ISSN 1198-6727 Mass-Balance Models of North-eastern Pacific Ecosystems Hsheries Centre Researeh Reports 1996 J.01wne 4 Nwnber 1 ISSN 1198-6727 ~~ --••11 Fisheries Centre ...... -~ · Research Reports 1996 Volume 4 Number 1

Mass-Balance Models of Northeastern Pacific Ecosystems

Fisheries Centre, University of British Columbia, Canada Dedicated to the memory ofProfessor P.A. Larkin

Mass - Balance Models of

North-eastern Pacific Ecosystems:

Proceedings ofa Workshop held at the Fisheries Centre

University ofBritish Columbia, Vancouver, B.C., Canada,

November 6-10, 1995

edited by

Daniel Pauly

and

Villy Christensen

with the assistance of

Nigel Haggan

Published by

The Fisheries Centre, University ofBritish Columbia

2204 Main Mall, Vancouver, B.C.

Canada, V6T lZ4

ISSN 1198-6727

1 Model construction was performed using the ABSTRACT well-documented Ecopath approach and software, previously applied to over eighty A one-week workshop was held at the aquatic ecosystems throughout the world, Fisheries Centre, UBC, from November 6­ and of which a pre-release Windows-based 10, 1996, during which ten invited version was applied during the workshop. participants, mainly from the scientific community in British Columbia, Alaska and This report documents the parametrization Washington, and Fisheries Centre faculty of the three above-me~tioned models and graduate students assembled the through short contributions authored by the elements required for preliminary mass­ participants, the construction and validation balance models of trophic fluxes in the of these (still) preliminary models, then Alaska Gyre, on the shelf off southern presents suggestions for their future British Columbia, and in the Strait of development and uses. Georgia.

Such mass-balance models were urgently required, as no systematic attempt had been made to verify that commonly-cited biomass, production and consumption rate estimates published for various critical marine groups in these three systems (e.g. salmon, marine mammals), were mutually compatible. The construction of these mass­ balance models not only allowed verification (or correction) of previously published flux and biomass estimates, but also identification of major gaps in knowledge, and cost-effective estimation of some of the previously unknown rates and biomasses required for assessment of marine carrying capacity in the Northeastern Pacific.

Each workshop participant covered a functional group and its associated fluxes: phytoplankton and primary production, zooplankton and secondary production, major fish species and their fisheries, marine mammals and birds and their food consumption.

3 short cut was invented by Dr. Jeffrey Director's Foreword Polovina from Hawai'i in the early 1980s. Ecopath is a straightforward ecosystem Researchers from academia, government modelling approach that balances the budget scientists and graduate students gathered at of biomass production and loss for each the UBC Fisheries Centre from November component by solving a set of simultaneous 6th-10th 1996, in order to construct the first linear equations. (The Ecopath approach is mass-balance models of three marine the only ecosystem model to obey the laws ecosystems in the North-east Pacific: the of thermodynamics!) Simple data on diet, Alaska Gyre, the shelf of southern British biomass, production and consumption to Columbia, and the Straight of Georgia. This biomass ratios is all that is required, so that report presents the result of the workshop as preliminary models can be quickly preliminary Ecopath models of these three constructed from data already published or systems. easily available.

Traditional fishery science conspicuously Ecopath's potential was recognized and was failed to take account of ecosystem championed at an early stage by Dr Daniel interactions and for many years the single Pauly at ICLARM, Manila. This resulted in species stock assessment reigned supreme. a series of workshops run world-wide, a Evaluation of the impact of harvesting on further development of the method, with predators and prey was often considered more rigorous mathematics by Dr Villy unnecessary: an interesting but purely Christensen from Denmark, and eventually, academic exercise. Many, including the late to the enhanced model published as Ecopath Peter Larkin himself, doubted this wisdom, II in 1992. Initially applied to a coral reef and a few pioneering souls dared to invent ecosystem by J.J. Polovina, over 80 Ecopath multispecies models or tried to model the models covering a wide range of marine and energy flow in whole ecosystems. But, by freshwater systems have now been emulating single species population constructed world-wide, most of them in the dynamics, multispecies fishery models developing world. It is something of a rapidly became over-parametrized and paradox that this new approach should be immensely data-hungry, while ecological used last in developed areas of the world ecosystem modellers went down ever more where the best data sets are available. The esoteric pathways, losing the mathematically power of the approach, which unites the ungifted on the way. Also, it soon became developed and developing world, lies in clear that whole-ecosystem models driven making cross-system comparisons possible. by primary production were swamped by flows to and from the microbial components The three Ecopath models in this report are ofthe ecosystem that are gigantic in relation preliminary and will doubtless need to be to exploited fish. They were also swamped refined with better and more precise data by massive uncertainty concerning the from the systems concerned. That can come nature and dynamics of those microbial later. But another, unexpected and exciting, flows. output arose from the workshop. This was the first version of Ecosim, a set of routines By concentrating on components of the added to Ecopath by Dr Carl Walters from system that were well described, a clever the Fisheries Centre in cooperation with

4 Dr Villy Christensen and Dr Daniel Pauly. Ecosim comprises a dynamic simulation of the effects of altering fishing mortality on selected components of the ecosystem. This technique, which allows the investigation of the impact of fishing on the whole ecosystem, will be formally published in 1997 in Reviews in Fish Biology & Fisheries. This new tool promises to be the first that could be practically employed in ecosystem management.

Mass-Balance Models of North-eastern Pacific Ecosystems is the seventh in a series of workshops sponsored by the UBC Fisheries Centre. The workshop series aims to focus on broad multidisciplinary problems in fisheries management, to provide an synoptic overview of the foundations and themes of current research, and attempts to identify profitable ways forward. Edited reports ofthe workshops are published in Fisheries Centre Research Reports and are distributed to all workshop participants. Further copies are available on request for a modest cost-recovery charge.

I thank UBC's Vice President of Research and the Department of Fisheries and Oceans Canada for sponsoring the workshop, and I am indebted to the late Dr Peter Larkin for sufficient shaking of their trees for the money.

Tony J. Pitcher

Professor ofFisheries

Director, UBC Fisheries Centre

5 The fourth and final reason relates to my FOREWORD current involvement in the North Pacific Universities Marine Mammal Research There are four reasons why Daniel Pauly Consortium which I helped set up three years might have asked me to make some ago. The Consortium is engaged in research introductory remarks for this workshop. that is at present concerned with the decline in abundance of Steller sea lions in the Gulf First, I helped raise some funds to support of Alaska, but in its initial conception, the the workshop. It wasn't as much as I would Consortium was encouraged to view the have liked to raise, nor as much as Daniel Steller sea lion question and other marine could have made good use of. But it was mammal issues in broad ecosystem contexts. enough to warrant a place on the program. I So what more natural than to encourage a would like to acknowledge a contribution of workshop on an Ecopath model of the Gulf $10,000 from the University of British ofAlaska? Columbia and a matching contribution from the Department of Fisheries and Oceans On behalf ofthe Consortium, I would like to Canada. add my welcome to that of Dr. Pitcher and wish you every success in the workshop. I Second, for several years now, Daniel has regret that I will not be able to participate to bent my ear about Ecopath, initially at the the extent that I had hoped but I will drop in International Center for Living Aquatic to the workshop during the week as I have Resources Management (ICLARM), and the opportunity. Onward and Upward! more recently here at UBC. I have read about Ecopath and formed some opinions about its usefulness for fisheries management, but I have to admit I have P.A. Larkin never really come to grips with this approach, and I suspect that others might North Pacific Universities Marine Mammal feel the same. To get that feel that comes Research Consortium with using a model, a workshop seemed a good idea. That is the second reason.

The third reason is that in recent years my attention has been much drawn to the Gulfof Alaska and the Eastern Bering Sea. This attraction stems in part from an interest that goes back to the days when I was involved in International North Pacific Fisheries Commission activities that has been rekindled by interactions with Lee Alverson (whose energies and whose capacities for generating enthusiasms are well known).

6 We take this opportunity, finally to thank all Preface and Acknowledgments those who made the workshop and this proceedings what they are now; Dr. Pitcher, The proceedings presented here pertain to a the Director of the Fisheries Centre, for his workshop held on November 6-10, 1995 support, Ms Rattana "Ying" Chuengpagdee, and, given our other commitments, we had for her superb organization of the event, Ms to limit our editorial interventions, as we Pamela Rosenbaum, for keeping us within wanted the proceedings to appear no more our budget, Ms Sandra Gayosa, of than a year after the event. ICLARM, for typing much of the first draft of this document, Mr Nigel Haggan for his Thus, what is presented here largely consists assistance in shaping it into a whole, the oftexts written by the workshop participants participants - including the Fisheries Centre before and/or during the workshop, the graduate students - for their enthusiasm, and major exceptions being contributions written UBC and DFO for funding the workshop in the Fall of 1996 by the editors, by J.1. and the publication ofthis report. Polovina and by N. Haggan, and presenting updates of recent developments concerning the construction and interpretation of Ecopath models, and their integration ill physical and cultural contexts. Daniel Pauly As a result, most readers should be able to find more pertinent references than we used, and and probably more accurate parameter Villy Christensen. values than were available during the workshop, and incorporated into what must Vancouver, October 1996 be viewed, therefore, as preliminary models. But then again, this was one of the main purposes ofthe exercise: to construct models that could serve as a basis for subsequent, more detailed work.

To encourage this, we will make available to anyone interested a copy of the Ecopath software, and of the files generated during the workshop (contact D. Pauly: [email protected], or V. Christensen: [email protected])

We dedicate these proceedings, modest as they are, to the memory of Professor P.A. Larkin, who passed away on July 10, 1996, and who made possible the workshop upon which these proceedings are based.

7 Table ofContents

ABSTRACT 3

DIRECTORS FOREWORD (T. PITCHER) 4

FOREWORD (P.A. LARKIN) 6

PREFACE AND ACKNOWLEDGMENTS (D. PAULYAND V. CHRISTENSEN) 7

TABLE OF CONTENTS 8

LIST OF EXHIBITS 10

INTRODUCTION 12

RATIONALE FOR MASS-BALANCE (TROPHIC) MODELS (D. PAULY AND V. CHRISTENSEN) 12 ALASKA GYRE MODEL 16

THE MARINE ECOSYSTEM OF THE GULF OF ALASKA (1.1. POLOVINA) 16 LOWER TROPHIC LEVELS (1. PURCELL) 16 CARNIVOROUS ZOOPLANKTON, JELLIES AND VELELLA (M. ARAI) 18 INITIAL ESTIMATES ON KRILL (A. JARRE-TEICHMANN) 20 SALMON IN THE OCEAN (L. HUATO) 21 MESOPELAGICS (1. PURCELL) 23 SHARKS (1. POLOVINA) 23 MISCELLANEOUS FISHES (P. LIVINGSTON) 24 MARINE MAMMALS (A. TRITES AND K. HEISE) 25 SEABIRDS IN THE ALASKAN GYRE (1. KELSON,. Y. WADA AND s. SPECKMANN) 31 BALANCING THE ALASKA GYRE MODEL (V. CHRISTENSEN) 32 SOUTHERN B.C. SHELF MODEL 37

THE AREA COVERED (D. PAULY) 37 EUPHAUSIIDS, CHAETOGNATHS AND HERBIVORES (1. PURCELL) 37 INVERTEBRATE BENTHOS (A. JARRE-TEICHMANN ANDS. GUENETTE) 38 CARNIVOROUS JELLIES (M. ARAI) 39 KRILL (A. JARRE-TEICHMANN) 40 SMALL PELAGICS (R. BUCKWSWORTH) 40 PACIFIC COD AND SABLEFISH (P. LIVINGSTON) 41 PACIFIC HALIBUT (1. VENIER) 43 SPINY DOGFISH (1. POLOVINA) 45 HAKE (T. PITCHER) 45 MARINE MAMMALS (A. TRITES AND K. HEISE) 51 SEABIRDS OF THE SOUTHERN B.c. SHELF (Y. WADA AND 1. KELSON) 55 FISHERIES CATCHES (E.A. BUCHARY) 57 BALANCING THE MODEL (1. VENIER) 59

8 STRAIT OF GEORGIA MODEL 63

SYSTEM DEF1NITION AND PRIMARY PRODUCTION (S. MACKINSON) 63 ZOOPlANKTON, INCL. JEU..IES (J. VENIER) 65 THE DEMERSAL FISH "Box" (J. VENIER ANDJ. KELSON) 65 THE PElAGIC "Box" (R. BUCKWOR11l) 67 MARINE MAMMALs AND BIRDS (Y. WADA) 69 FISHERIES HARVEST INTHE STRAIT OF GEORGIA (E.A. BUCHARY) 73 BAlANCING THE STRAIT OF GEORGIA MODEL (J. VENIER) 74 GENERAL DISCUSSION••••••••••••••••••••••••.•••••••••••.•••••••••••••••••••••••••••••••••••••.•••••••.••••••••.•••.•••...... 81

A ROAD TEST OF ECOPATH (R. BUCKWOR11l) 81 SUGGESTED IMPROVEMENTS FOR ECOPATH MODELING (C. WALTERS) 81 EXPLORING ECOSYSTEM RESPONSES TO ENVIRONMENTAL VARIATION (J.J. POLOVINA) 82 INTEGRATION OF LOCALENVIRONMENfAL KNOWLEDGE (N. HAGGAN) 88 UPDATESONEcoPATHDEVELOPMENf ANDAPPUCATIONS (D. PAULY AND V. CHRISTENSEN) 89 REFERENCES 92

APPENDIX 1 116

TABLES A-O 116 APPENDIX 2 ••••••••••••••••••••.••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••.••••.••••••••••••••••••••••••••••••••••.••••.. 128

WORKSHOP SCHEDULE 128 APPENDIX 3 130

LIST OF PARTICIPANTS 130

9 List ofexhibits p. Box 1 Basic equations, assumptions and parameters ofthe Ecopath approach .l5 Fig. 1 Flow chart oftrophic interactions in the Alaska gyre system .34 Fig. 2 Mixed trophic impacts in the Alaska gyre system 35 Fig. 3 Flow chart oftrophic interactions in the shelfsystem offsouthern B.C 62 Fig. 4 Flow chart oftrophic interactions in the Strait ofGeorgia 78 Fig. 5 Mixed trophic impacts in the Strait ofGeorgia ecosystem 79 Fig. 6 Trophic pyramids for three ecosystems ofthe North-eastern Pacific 89 Table 1 Length-weight data for Velella velella 19 Table 2 Freshwater, estuarine and ocean residence times of salmonids 21 Table 3 Estuarine and oceanic mortality rates ofPacific salmon .21 Table 4 Numbers and biomass of sockeye, pink and chum in the Alaska gyre 22 Table 5 British Columbia chinook and coho catches, statistical areas 11 - 27 .22 Table 6 Stomach contents ofsalmonids in the Subarctic Northern Pacific 23 Table 7 Diet composition ofchinook salmon offVancouver Island .23 Table 8 Biological and population statistics ofmarine birds, Alaska gyre 31 Table 9 Diet composition ofmarine birds in the Alaska Gyre 32 Table 10 Basic parameters for Alaska gyre model 33 Table 11 Benthic groups and their PIB ratios, southern shelfofB.C 38 Table 12 Suggested PIB, GE and EE values for benthic groups, southern B.C 39 Table 13 Diet and conversion efficiencies for benthos ofthe B.C. shelL 39 Table 14 Pacific hake biomass, B.C. coast .48 Table 15 Summer diet composition ofhake by size, U.S. coast.. 50 Table 16 Summer diet composition ofhake, Vancouver Island 50 Table 17 Summer and winter seabird populations and food consumption 56 Table 18 Diet composition of seabirds, southern B.C. shelf 57 2 Table 19 Mean landings per km , southern B.C. shelf. 58 Table 20 Data for estimation of discarding by B.c. trawlers 59 Table 21 Basic parameters, southern B.C. shelfmodel 61 Table 22 Standing crop ofLaminaria spp at various N. Atlantic sites 64 Table 23 Summary statistics: primary production in the Strait ofGeorgia 64 Table 24 Biomass, PIB and Q/B ofzooplankton in the Strait ofGeorgia 65 Table 25 Definition ofdepth strata for the Strait of Georgia 65 Table 26 Bottom types in the intertidal areas ofthe Strait ofGeorgia 66 Table 27 Estimates ofbiomass, PIB, Q/B, P/Q and EE for macrobenthos 66 Table 28 Diet composition ofherbivorousldetritivorous benthos and crabs 67 Table 29 Biomass, PIB and related statistics ofdemersal fishes, Strait of Georgia 67 Table 30 Ecological parameters: pelagics ofthe Strait of Georgia ecosystem 68 Table 31 Diet composititon ofpelagic box, Strait ofGeorgia 69 Table 32 Estimates ofsalmon abundance, Strait ofGeorgia 69 Table 33 Body weight and related statistics of marine mammals, Strait of Georgia 71 Table 34 Body weight and related statistics ofmarine birds, Strait ofGeorgia 73 Table 35 Commercial fisheries harvests, Strait of Georgia (Summer) 74 Table 36 Basic parameters, Strait ofGeorgia model 78

10 Appendix 1: Table A Biological and population statistics ofmammals, Alaska gyre (Summer) 116 Table B Biological and population statistics ofmammals, Alaska gyre (Winter) .116 Table C Diet composition ofmarine mammals in the Alaska gyre (Summer) 117 Table D Diet composition of marine mammals in the Alaska gyre (Winter) 117 Table E Diet matrix for Alaska gyre ecosystem model .118 Table F Population statistics ofmarine mammals, southern shelfofB.C. (Summer) .119 Table G Population statistics ofmarine mammals, southern shelfofB.C. (Winter) 119 Table H Diet composition ofmarine mammals, southern shelfofB.C. (Summer) 120 Table I Diet composition of marine mammals, southern shelfofB.c. (Winter) 121 Table J Diet matrix for southern B.c. shelfecosystem model .122 Table K Mean seasonal landings ofthe fisheries, southern B.C. shelf.. 123 Table L Composition ofthe macrobenthos ofthe Strait ofGeorgia, by depth .124 Table M Diet composition ofmarine mammals .125 Table N Diet composition ofmarine birds 126 Table 0 Diet matrix for Strait ofGeorgia ecosystem model 127

11 between them. It thus provides an INTRODUCTION independent way to validate published estimates ofbiomasses and/or fluxes.

(Daniel Pauly and Villy Christensen) However, most marine and fisheries scientists have to date preferred to work on single species, or on interactions among few species, rather than to model entire Rationale For Mass-Balance ecosystems. There are many reasons for this, (Trophic) Models but the most prevalent is that "ecosystems modeling" is perceived as a specialized and Over the last decades, there have been long-term activity, leading to large, massive changes of the oceanic regime in unwieldy products, perhaps best illustrated the North Pacific. These changes have by the Bering Sea model alluded to above. resulted in increase of some major resource Practical demonstrations of the power and species, notably several species of salmon versatility ofthe Ecopath approach may help off British Columbia (B.C.) and Alaska to shift emphasis to whole-system attributes, (Beamish and Bouillon 1993), and hake off increasingly required in an age of global B.C. (Pitcher, this vol.). Decreases in other change. species, e.g., lobsters in Hawai'i have been attributed to the same climatic changes (Polovina et al. 1994). The Ecopath approach and software

Oceanographers, fisheries scientists and The Ecopath approach is based on the work marine biologists have been tracking these of J.J. Polovina (1984), of the US National changes, sometimes together, more often Marine Fisheries Service (Honolulu separately. Their common goal has been to Laboratory). The present authors, both then predict future oceanographic regimes and, at the International Center for Living based thereon, the likely futures of major Aquatic Resources Management (lCLARM) resources species. This research has yielded in Manila, Philippines, developed and excellent results. adapted Polovina's work to a user-friendly and versatile software package for personal However, the "trophic context" of key computers. This software allows for rapid species in north Pacific ecosystems has been construction and verification of mass­ largely neglected, at least since T. Laevastu balance models of ecosystems (Christensen and colleagues' heroic attempt at a spatially­ and Pauly 1992a, 1992b). structured simulation model of the entire Bering Sea (see Laevastu and Larkins 1981). The key steps to construct a model are to: Trophic context, as defined here encompasses food consumption and i) Identify the area and period for which a requirements of key species, and their model is to be constructed; contribution to the diet of predators. Dealing with the trophic context of resource species ii) Define the functional groups (i.e., provides powerful insights for estimation of "boxes") to be included; biomasses and of the fluxes that occur

12 iii)Enter a diet matrix defining all trophic British Columbia, and the states of Alaska linkages by expressing the fraction that and Washington, US, Fisheries Centre each "box" in the model represents in the faculty and graduate students. Dr. Polovina diet ofits consumers; was invited for the workshop to benefit from his pioneering work on Ecopath, and more iv)Enter the food consumption, recently, on regime shifts in the North production/biomass ratio and/or biomass, Pacific gyre (Polovina et al. 1994, 1995). and fisheries catches, ifany, for each box; Identification of key scientists was helped v) Balance the model, i.e., modify entries by discussions ofthe mass-balance approach (iii & iv) until input = output for each at the March 1995 UBC Fisheries Centre box; workshop on "Impact of Changes in North Pacific Oceanographic Regimes on Coastal vi) Compare model outputs (network Fisheries." Aside from the travel characteristics, estimated trophic levels arrangements, preparatory work included the and other features of each box) with briefing prospective participants on the data estimates for the same area during to be assembled, the compilation of a another period, or with outputs of the preliminary database and assembling same model type from other, similar relevant publications. Also, the Ecopath areas, etc. "module" taught (by D.P.) in September­ October 1995 at the Fisheries Centre as part of"Fish 504"used the Strait ofGeorgia as an These steps are simple provided that basic example, thus preparing the students to parameters are available. Numerous well­ participate in the workshop, which became documented Ecopath applications to aquatic an element oftheir course. ecosystems already exist, ranging from The first day of the workshop proper was aquaculture ponds and flooded rice paddies devoted to a formal review of the concepts to shelf systems (see Pauly and Christensen behind the Ecopath approach (Box 1) and 1993, and contributions in Christensen and informal presentations and discussions on its Pauly 1993). Other models exist which various features, both positive and negative. represent the North Sea (Christensen 1995), The following multi-authored text presents and George's Bank. Indeed, the latter model the following: was constructed by a group of Canadian scientists (from DFO and other institutions) • Estimated "summer" biomasses of during a workshop run by V. Christensen, the major components in three similar to the one reported upon here, and marine ecosystems: the Alaska Gyre, held in St. John's in late 1993. on the shelf of Southern B.c., and in the Strait Georgia; The workshop and its outputs • Mass-balance models describing the The workshop was held November 6-10, trophic flows in these three 1995, at the Fisheries Centre, UBC. There ecosystems, and largely pertaining to were some twenty-four participants: mainly the late 1980s; members of the scientific community in

13 • Parameter estimates for the biomasses of key species in the three construction of "winter" models for model areas . these same areas, and hence for studying seasonal cycles; We hope these results will be seen as useful by the scientific community in the Pacific • A basis for comparisons with similar Northwest, and contribute to a renaissance ecosystems in other parts of the of studies linking ecological models and world; and, most importantly: applied fisheries work. We also hope that the overview of ecosystems we present here • A basis for formulating hypotheses will also lead to more consideration of local about the likely effect of future knowledge on their interrelationships of changes of oceanographic and/or their component species (see Haggan, this fisheries regimes on food webs and vol.).

14 Box 1 Basic equations, assumptions and parameters of the Ecopath approach.

The mass-balance modelling approach used in this workshop combines an approach by Polovina and Ow (1983) and Polovina (1984, 1985) for estimation of biomass and food consumption of the various elements (species or groups of species) of an aquatic ecosystem (the original "ECOPATH") with an approach proposed by Ulanowicz (1986) for analysis of flows between the elements of ecosystems The result of this synthesis was initially implemented as a DOS software called "ECOPATH II", documented in Christensen and Pauly (1992a, 1992b), and more recently in form of a Windows software, Ecopath 3.0 (Christensen and Pauly 1995, 1996). Unless noted otherwise the word "Ecopath" refers to the latter, Windows version.

The ecosystem is modeled using a set of simultaneous linear equations (one for each group i in the system), Le.

Production by (i) - all predation on (i) - nonpredation losses of(i) - export of(i)= 0, for all (i).

This can also be put as

Pj-M2i - Pj(I-EEj)- EX j= 0 I)

where Pj is the production of (i), M2 j is the total predation mortality of(i), EEj is the ecotrophic efficiency of (i) or

the proportion of the production that is either exported or predated upon, (1-EEj) is the "other mortality", and EXj is the export of(i).

Equation (1) can be re-expressed as

Bj*P/B j - LP/Q/B/DCij-P/Bj*Bj(I-EEj)-EXj =0

or

..2)

where Bj is the biomass of (i), P/Bj is the production/biomass ratio, Q/Bj is the consumption/biomass ratio and DCij is the fraction of prey (i) in the average diet of predator 0).

Based on (2), for a system with n groups, n linear equations can be given in explicit terms:

.B1P/B\EE 1 - BIQ/BIDCII-BzQ/BzDCzl - ...-B"Q/B"DC"I - EX 1 = 0

.BzP/BzEEz- B1Q/B)DC 1Z - BzQ/BzDC zz - ...-B"Q/B"DC"z - EXz = 0

B"P/B"EE" - B1Q/BPC1"- BzQ/BzDCz" - ...-B"Q/B"DC"" - EX" = 0 This system ofsimultaneous linear equations can be solved through matrix inversion. In Ecopath, this is done using the generalized inverse method described by MacKay (1981), which has features making it generally more versatile than standard inverse methods.

Thus, if the set of equations is overdetermined (more equations than unknowns) and the equations are not consistent with each other, the generalized inverse method provides least squares estimates which minimize the discrepancies. If, on the other hand, the system is undetermined (more unknowns than equations), an answer that is consistent with the data (although not unique) will still be output.

Generally only one of the parameters Bj, P/Bj, Q/B;, or EEj may be unknown for any group L In special cases, however, Q/Bj may be unknown in addition to one of the other parameters (Christensen and Pauly I992b). Exports (e.g., fisheries catches) and diet compositions are always required for all groups.

A box (or "state variable") in an Ecopath model may be a group of (ecologically) related species, i.e., a functional group, a single species, or a single size/age group of a given species.

15 1946-76 when the Aleutian Low was weaker ALASKA GYRE MODEL than average.

As the late 1970s represent a transition The Marine Ecosystem of the Gulf period from low to high carrying capacity, ofAlaska our trophic model ofthe Alaska Gyre system should focus on the 1980s, and describe the ecosystem characteristics of this latter (Jeffrey Polovina) period.

The Alaska Gyre is an extremely important part of the North Pacific: it is the ecosystem Lower Trophic Levels where most salmon from the Pacific Northwest accumulate the energy that enables them to swim back to the stretch of (Jenny Purcell) the river or brook where they once hatched. The data presented here originate mainly Compared to earlier decades, there is a from the SUPER (SUbarctic Pacific significant body of biological evidence to Ecosystem Research) project, (see Miller et suggest that the carrying capacity ofthe Gulf al. 1988) and were sampled during May­ ofAlaska has increased since the late 1970s. June and August 1984, 1987 and 1988 at 0 Changes have occurred at several trophic two locations - Station P (500 N; 145 W) 0 0 levels from zooplankton to at least salmon. and Station R (53 N; 145 W). All standing This is possibly due to changes in the stock estimates, integrated to a depth of intensity of the Aleutian Low Pressure 80m, and initially expressed in terms of -2 d -I System (McFarlane and Beamish 1992; mgC'm . ay , were converted to g wet 2 l Beamish and Bouillon 1993; Brodeur and weight m- 'year- through a wet weight : Ware 1992; Polovina et al. 1995). carbon ratio of 10: 1.

A comparison of zooplankton abundance in the Gulf of Alaska between 1956-1962 and Bacteria 1980-1989 showed a doubling of summer Biomass and production data for bacteria zooplankton biomass, pelagic fish, and squid were taken from Kirchman et al. (1993; abundance in the latter period (Brodeur and Tables 1 and 2, respectively): Ware 1992). Trends in North Pacific salmon production follow changes in the Aleutian 2 • Biomass 1.1 gC m- = 11 g wet weight Low Pressure Index from 1925-1989 m-2,• (Beamish and Bouillon 1993). Above 2 average North Pacific salmon catches • Production 55.6 mg C m- day"l = 203 g occurred during 1925-1945 and 1977-1989, wet weight m-2 year-I; when the Aleutian Low Pressure System l was more intense than average. Below • PIB = 18.45 year- ; average salmon catches occurred during

16 • Carbon consumption (taken for 0 - 80 m Small herbivores from Table 1 in Simon et al. 1993). Density estimates for eight specIes of Thus, we have 75.5 mgC m-2 dai1 = 276 g copepods were adapted from Table 1 in wet weig. ht m-2 year,-I or Q/B = 25 year,-I Dagg (1993), based on data collected with a which is low such small organisms. MOCNESS net (243 ~m mesh). Dagg et al. (1989) report biomasses of 85 ~gC for Neocalanus plumchrus and 624 ~gC for N Phytoplankton cristatus (see also Frost 1987). Densities were converted to carbon using 85 ~gC, Chlorophyll a biomass (from Table 2 in assuming that most copepods were small. Frost 1993) was converted to carbon using Herbivore biomass is about 15 g m-2 in the to the ratio of 60: 1. Chlorophyll biomass is summer and about 4.5 g m-2 in the winter rather constant over the year (Miller et al. (Richard Brodeur, pers. comm.). Thus, we 1991). Thus we have: have: 2 • Biomass 25.3 mgChl. 'm- = 15.2 g wet 3 . h -2 • Biomass 535.6 copepods m- (80 m weig t'm ; 2 depth) = 29,272 copepods m- , or = 25 g 2 wet weight m-2 ; • Production 728.4 mg C m- da/ = 2659 2 l g wet weight'm- year- ; • Copepod ingestion of phytoplankton was estimated from Figure 12 in Dagg (1993), • P/B = 175 year-I. as: 943.8 rgChl. m-2 dail = 2067 g wet weight m- year-I. Futher, data may be found in Welschmeyer et al. (1993). Clearance rates of protozoans by N plumchrus and N cristatus were averaged Microzooplankton from data in Gifford (1993). All copepods were assumed to clear protozoans at the Biomass and ingestion rates of same rate (this may be an overestimate, see microzooplankton are from Strom et al. Landry et al. 1993). Thus: (1993; Tables 1 and 2, respectively): Ingestion of protozoans 1.8 ~gC copepod-I 2 • Biomass 170.4 mgC m- = 1.7 g wet hour-I = 192 g wet weighf2 year-I, and Q/B 2 weight m- ; is thus 90.4 year-I.

2 l • Ingestion 98 mgC m- dai = 358 g wet Adding phytoplankton and protozoans gives 2 weight m- year-I; a total copepod ingestion rate of 2,259 g wet weight'm-2 year-I, and the diet based on the • Q/B = 0.6 per day or 210 year-I. ratio of those consumption rates is 91.5% phytoplankton and 8.5% (bacteria-rich) The diet consists of 75% phytoplankton and detritus. 25% bacteria and other microbes.

17 Salps andgelatinous herbivores/omnivores in Terazaki and Miller (1986) show a later summer maximum approximately double The major species of salp is Cyclosalpa that for May. bakeri, and the following, adapted from Purcell and Madin (1991) and Madin and Terazaki and Miller (1986) estimated Purcell (unpublished data), refer only to that generation times of 6-10 months; thus, with species, though Salpa fusiformis also occurs three spawning periods at Station P, P/B is t in the area. Salps were not present in the at least 1.5 year- , and we assume it here to May cruises of SUPER, but were in August. be twice as high. Overall, salp may occur only 3-4 months per year in the Alaska Gyre. Cyclosalpa bakeri The diet consists mainly of small herbi­ is extremely delicate, and its biomass and vorous zooplankton (see Sullivan 1980). related parameter are unlikely to have been estimated in any other studies. The values Carnivorous Jellies are:

2 The numbers ofAglantha were derived from • Biomass 804 mgC m- = 8 g wet weight -2 data in Forbes et al (1988) for May 1984. m ., This value is probably an underestimate as it 2 1 is approximately one-third of the value in • Total ingestion 66.3 mgC m- dai = 242 g wet welg. ht·m -2 year;-I and Arai and Fulton (1973) for 1971. The estimate also does not include • Q/B = 30 year-I. siphonophores, which, in 1971, were approximately two-thirds as abundant as The diet consists of 53% phytoplankton and Aglantha (Marlow and Miller 1974). 40% microzooplankton and 7% detritus. Numbers were converted to wet weight Further data on the metabolism and growth using a value of 83.2 mg/animal (Ikeda -2 of Cyclosalpa bakeri may be found in 1972). Thus, biomass = 9.1 g·m , with Madin and Purcell (1992), and in Cooney winter populations similar to those in May (1987, 1988) for zooplankton in general. (Arai and Fulton 1973).

On the high sea, there is one more Carnivorous Zooplankton, Jellies generation of Aglantha than inshore. P/B is and Velella therefore = 1+ predation rate, and probably t ranges from 2 - 4 year- • (Mary Arai) A first estimate of ingestion rate, derived from Purcell and Grover (1990) is QIB = Carnivorous Zooplankton at Station P 110 year-I, which appears to be very high, perhaps excessively so. Therefore, Q/B is Populations of chaetognaths at Station P estimated from an assumed food conversion were estimated from Forbes et al. (1988). efficiency of 0.3. The diet is assumed to Standing stock in May was 2,078 consist ofsmall herbivorous zooplankton. 2 individuals·m- . The standing stock graphed

18 QJ Table 1 Data for establishing a length-weight relationship in J.': velella •

Length Weight Length Weight Length Weight 8 50 19 360 25 710 11 110 21 420 26 490 15 200 22 480 27 880 15 210 22 520 28 880 16 280 22 670 28 990 17 340 23 600 29 840 18 270 23 650 31 940 18 310 23 660 32 1400 18 360 24 810 36 1590 18 370 - - - - a) based on hydrOIds sampled In May 1986, stored In 5% formalin, , and measured (mm and mg wet weight) in October 1996.

Gyre system, nor for the southern B.c. shelf, By the Wind Sailor Velella velella where V velella also forms large aggregations offthe B.C. coast. Velella velella is a hydroid which form floats with sails, hence its common name, For the B.C. coast, population numbers as "By the Wind Sailor". Like coral, V velella length categories for spring populations in contain symbiotic algae (Hovasse 1923; neuston tows are given in Arai et al. (1993). Taylor 1971; Holland and Carre 1974), but Pertinent stations for the 30-300 m shelf are they also feed on euphausiid and fish eggs LB2-11 and LC 2-8. A length-weight (Bieri 1961). The ratio of food intake vs. relationship being necessary to analyze these energy supply from the algal symbionts is data, but none having so far been published, not known. Growth to float length of 10 cm 28 preserved specimens of V velella were occurs from January to April; along the measured and weighted for the purpose of California coast, a second generation occurs deriving such relationship (Table 1). from July to September (1977). From these, the relationship The northern limits of V velella distribution 2 26 roughly correspond to the convergence of W = 0.48·L . subarctic waters with the mixed waters of the Alaska current (Savilov 1961), and was established, using a linear regression of hence the distribution of V velella swarms is the 10g(W) against the 10g(L) values, whose very irregular in space and time (see also correlation is r = 0.981. This relationship Bieri 1977, and Arai 1992). For example, in may be used to convert observed lengths of 1957, seven swarms appeared at ocean V velella into weights. However, the narrow weather station "PAPA" (50° N, 145°W) in range of sizes used to derive the relationship February, May and October. Thus, biomass must be considered, as well as a weight loss estimates are not provided for the Alaska following preservation, ofperhaps 30-50 %.

19 Initial Estimates on Krill detritus. However, a quantitative breakdown was not available. Euphausiids are known to be predominantly filter feeders, although (Astrid Jarre-Teichmann) some hunting has been observed (Mauchline and Fisher 1969; Lasker 1966). A diet Euphausia pacifica, Thysanoessa longipes, composition of 85% phytoplankton, 5% and T are the most important inermis zooplankton, and 10% detritus may be euphausiid (krill) species in the central assumed. northeast Pacific, (Mauchline and Fisher 1969). T inspinata occurs further south than A population of E. pacifica uses about 9% T longipes, and the northern boundary of its of the assimilated carbon for somatic distribution lies south of 50° N. T raschi is production over the entire life of its more common on the shelfthan offshore. constituent individuals; mature animals use another 9% for gonadal production, whereas T longipes and T inermis live for 2-3 years the somatic production ofjuveniles can be as (Mauchline 1980), while E. pacifica lives high as 30% (Lasker 1966). Therefore, the for at least 2 years (Tanasichuk 1995). Most net efficiency of E. pacifica was assumed at species mature at about one year of age 20%, in line with the estimates of Lasker (Mauchline 1980). Lindley (1980) estimated Z = 0.8 year-! for 2-year old T inermis. (1966) that 62-87% of the assimilated carbon is respired. Lasker also estimated an Total mortality (Z) estimates for E. pacifica assimilation coefficient of 80% for E. range from Z = 0.6-1.9 year-) on the shelf of pacifica fed with nauplii, but, as the B.C. (Ron Tanasichuk, Nanaimo, pers. assimilation efficiency of omnivores is comm.) to Z = 3.0 year-\ in the California generally lower than that of carnivores Current (Mullin 1969) and Z = 8.7 year-! off Oregon (Mauchline 1980). Z = 3.0 year-! (Welsh 1968), an assimilation coefficient of 70% is suggested, close to the estimates of was chosen as an initial estimate for the other cold-water zooplankton of 65% present modelling study. This may tum out (Schnack et al. 1985). This leads to a gross to be conservative if juvenile stages of krill efficiency of 12.6%. It should be noted that are to be explicitly included in this box. this value is considerably higher than the E. pacifica grow rapidly during summer estimate of 5.4% for Euphausia superba (Lasker 1966), but as in other organisms, based on the energetic model of Clarke and growth stagnates for several months during Morris (1984) and the annual P/B ratio of winter (Ron Tanasichuk, Nanaimo, pers. Siegel (1986). comm). Consequently, if a typical summer situation is to be represented on an annual In the absence of biomass estimates for the basis, the initial PIB ratio should be set at 5- Alaska gyre, krill biomass can be computed, 6 year.-I based on an ecotrophic efficiency of 95%, which emphasizes the role ofkrill as food in In general, the diet of both T longipes and the ecosystem. Also, the total zooplankton E. pacifica consists of detritus, diatoms, biomass estimate of 34.5 g m-2 at an oceanic dinoflagellates, tintinnids, chaetognaths, station in the northeastern Pacific "transition larvae of echinoderms, amphipods, and zone" in the late 1980s (Richard Brodeur, crustaceans (Mauchline 1980), i.e., of pers. comm.) may serve as a guiding value phytoplankton, small zooplankton and for balancing the summer model.

20 Salmon in the Ocean feeding. All mortality can be attributed to predation as there is no US or Canadian commercial fishing during their oceanic (Leonardo Huato) phase. Coho and chinook remain in coastal waters. Table 3 presents instantaneous The northeastern Pacific ocean supports mortality estimates for the estuarine and numerous stocks of six anadromous oceanic phases. salmonids species: coho, Oncorhynchus kisutch; chinook, 0. tshawytscha; sockeye, Biomass or densities are not reported in the 0. nerka; steelhead, 0. mykiss; chum, 0. literature. For steelhead, biomass was set at 2 keta; and pink, 0. gorbuscha (Healy 1993; 0.1 t'km- , and P/B at 1 year-I. For the other Ignell and Murphy 1993). Table 2 presents species (except chum), biomass was freshwater, estuarine and ocean residence estimated using total catch and total times for 5 ofthese 6 species: mortality (Z; year-\ and harvest rate (annual catch in numbers/population size), Sockeye, pink and chum salmon migrate as follows: into oceanic waters. On their return, they pass throughout the coastal domain without • Current year run size = total catch / Table 2 Freshwater, estuarine and ocean harvest rate; and residence times for salmonids (Pearcy 1992). • Size of the cohort in the previous Species Freshwater Estuary Ocean z year = current year run·e .

Coho 0- 4 years days 0.5 - 1.5 years For chum, the catch was assumed to Chinook 0- 2 years days 0.5 - 6 years be composed of 30% of individuals of age 4 and 70% of age 3 individuals. Sockeye 0- 2 years days 1 - 5 years Thus, the total number of age 4 is Chum days-weeks weeks 2 - 4 years given as:

Pink days-weeks days 1.6 years • N4 = 0.3·total catch/harvest rate;

Table 3 Estuarine and oceanic mortality rates (Z; • N3 = (0.7·total catch/harvest rate) year-I) for salmon, by stage (based on Bradford Z 1995, and Ricker 1976). + N4 ·e ;

Species Smolt-Adult Adult (Estuaries) (Ocean) z Coho 2.32 ± 0.06 1.32 • N 1 = N2 ·e (Table 4). Chinook - 0.42 Chum 4.92 ± 0.12 1.64 Total catches were obtained by pooling US and Canadian catches in Sockeye 2.78 ± 0.07 0.92 numbers, then converting to biomass Pink 3.68 ± 0.12 2.45 using mean body weights at age.

21 Table 4 Numbers and biomass of sockey.e, pink and chum salmon in the Alaska Gyre (U =barvest rate; Z =total mortality; year-\

Sockeye Pink Chum U= 0.6 U=0.7 U = 0.55 Z= 0.92 Z = 2.45 Z = 1.65 Catch = 50 million Catch = 72 million Catch = 12 million Numbers Weight Biomass Numbers Weight Biomass Numbers Weight Biomass 6 6 at age (10 ) (kg) (t) at age(10 ) (kg) (t) at age(l06) (kg) (t) 83 3 250 101 1.7 172 6.5 6 3.9 209 0.72 151 588 0.3 176 49 3 147 525 0.11 57.7 253 .Il .28 Density (t·km-2) = 0.109 Density (t·km-2) = 0.083 Density (t·km-2) = 0.051

The exploitation rates used here were kindly sockeye, pink and chum salmon. provided by Carl Walters (pers. comm.). Sockeye catches are from the early 1980s As stated above, chinook and coho do not and stem from Burgner (1991); pink catches have an oceanic phase, i.e., they are only are as reported by Heard (1991) for the years coastal residents. Catches for this species are 1980-1989; chum catches stem from Salo for DFO statistical areas 11-27 (Table 4). (1991) and pertain to the early 1980s. Diet composition varies with location, The area of the Northeast Pacific (i.e., season and length of the fish, and Table 6 Alaska Gyre) over which the biomass was gives a first example. The "other" category assumed to be distributed was 4,205,000 in that table includes identified preys for krn2, and 30,000 krn2 for the shelf and shelf pink, sockeye, coho and steelhead, and edge of southern S.c. (from the Southern tip unidentified preys for chum. As chum, with of Vancouver Island up to the Southern tip their large stomachs, are known to be of the Queen Charlotte Islands, and from specialized to feed on coelenterates (Arai depths of 30 to 300 m; see Pauly, this vol.). 1988), it may be assumed that the large Table 4 presents catch numbers used in the fraction of their unidentified food consisted calculation and estimated biomasses of ofjellyfish.

Table 5 British Columbia chinook and coho catches from statistical areas 11 - 27 from 1980 to 1989 (U = harvest rate; Z = total mortality; year-); source: DFO, D.C. Catch Statistics Reports).

Coho Chinook U = 0.4 U =0.4 Z- 2.32 Z = 0.42 Catch - 1,666,667 Catch = 100,000 Weight Numbers Biomass Numbers Weight Biomass 3 (kg) at age(l0 ) (t) at age(l06) (kg) (t) 20 250. 5000 4.17 3 12,500 to 380 3805 42.4 1 42,398 2 579 1158 ------0.1 881 88 ------Density (t·km-2) = 1.830 Density (t·km-2) = 0.335 22 Table 6 Diet composition (% weight) of Table 7 Diet composition of chinook pink, sockeye, chum, coho and steelhead salmon, in % volume (based on 30 g in the Subarctic Northern Pacific (from fish caught off Vancouver Island LeBrasseur, 1966). Healey 1991).

Pred.\Preys Amphipods Squid Fish Other Prey Species 0/0 Pink 15 75 10 - Herring 30 Sockeye 20 75 2 3 Sand Lance 18 Chum I 2 - 97 Pilchard 7 Coho - 100 - - Anchovy 4 Steelhead 2 95 I 2 Rockfish 2 Other fish 9 Euphausiids 25 Table 7 presents another example. Squids 4 Other invertebrates I Healey (1978) also reported on the food and feeding habits of coho salmon in the Strait of Georgia, based on fish with fork lengths (Gj0saeter and Kawaguchi 1980) using an of 11.6 - 28.1 em. The stomach contents M/K ratio of2.0. were composed of herring, sand lance and Important dietary items include copepods, unidentified fish remains (32%, by volume); euphausiids, ostracods, amphipods, and amphipods (33%) and crab megalops (26%). small decapods (Gj0saeter and Kawaguchi Stomach content as percentage of body 1980). The food consumption of Diaphus weight varied between 0.4% and 1.5%. taaningi was estimated at 0.8% body weight l day" I, or 3 x body weight year- (Baird et al Mesopelagics 1975).

(Jenny Purcell) Sharks

Mesopelagic fishes in the Gulf of Alaska are a multispecies group, characteristically (Jeffrey Polovina) found between 150 - 500 m during the day and between 150 m and the surface at night The principal oceanic sharks in the Gulf of (Gj0saeter and Kawaguchi 1980). Alaska are the salmon shark (Lamna Stenobrachius leucopsarus is the dominant ditropis) and the blue shark (Prionace species in trawl catches. Other frequently glauca). Salmon sharks are year-round caught species include Diaphus theta, residents. An estimate of natural mortality Tarletonbeania crenularis, and T macropus for a close relative of the salmon shark, the (Gj0saeter and Kawaguchi 1980). porbeagle (Lamna nasus) is 0.18 year-\ (Aasen 1963). Blue sharks are summer The density of the mesopelagic group in the visitors which migrate north from transition Gulf of Alaska is estimated at 4.5 t'km-2 zone waters (Brodeur 1988). Natural (Gj0saeter and Kawaguchi 1980). An mortality range for blue sharks is 0.18-0.24 estimate of P/B for the group, of 0.7 year-\, year-\ (Nakano and Watanabe 1992), which is obtained from the von Bertalanffy growth suggests an estimate of P/B of about 0.2 parameter K = 0.34 year-\ for S. leucopsarus year-\ for both species.

23 Salmon sharks appear to feed primarily on the 10° C isotherm (Trumble 1973). Small coho, sockeye, pink, and chum salmon but pomfret «30cm FL), are not found north of may also consume mesopelagics and other 44° N in July (Pearcy et a1.1993). Pomfret pelagic fishes (Brodeur 1988; Compagno are usually the most abundant non-salmonid 1984). Blue sharks feed on squid, fish in the subarctic region of the North mesopelagics, saury, and pomfret (Brodeur Pacific. 1988; Compagno 1984). While food consumption is not known for the salmon or The diet of pomfret, as summarized by blue sharks, for the mako shark it has been Brodeur (1988), consists mainly of estimated at 3% body weight per day or cephalopods and fish which, in the majority about 10 times body weight year-I (Stillwell of studies, comprised over 50% of the diet and Kohler 1982). by weight. Other preys, contributing from 11 to 49% of the diet by weight, were The densities of salmon shark and blue euphausiids, amphipods, and decapods. sharks in the Gulf of Alaska are not known. Pearcy et al. (1993) found that squids made As a lower bound, the density of sharks, up over 75% of the prey volume. Usually, principally blue and salmon, caught as by­ small Gonatus spp. (DML <60mm) were catch in the squid driftnet fishery in 1990, is found in pomfret taken in the northern Gulf estimated at 0.05t·km-2 (Bonfil 1994). Blue of Alaska. Myctophids were the most sharks are absent from the Gulf of Alaska in frequently occurring fish in the diet of winter. Hence, based on ratios of salmon pomfret in the Gulf of Alaska north of 45° shark to blue shark in the bycatch, the winter N, but saury were also found. Here, diet shark biomass is probably only one fourth of percentages by weight were estimated using the summer value (Bonfil, 1994). information in Pearcy et al. (1993), leading to the following: 75% small squid; 3% amphipods; 8% mesopelagics; 4% saury; Miscellaneous fishes 5% carnivorous zooplankton; 5% large herbivorous zooplankton.

(P. Livingston) Population-weighted food consumption (Q/B) was estimated as 4.28 year-I using Pomfret (Brama japonica) equation (3) of Pauly et al. (1993b), an asymptotic length (L~; TL) of 61 cm based The pomfret is an epipelagic fish species on Eschmeyer et al (1983), and data in occurring in the subarctic zone of the North Shimazaki (1989), from which an Pacific in summer, but apparently not in asymptotic weight (W~) of 3859 g at a winter (Brodeur 1988). Relative abundance temperature.of 12 °C was obtained. in the Alaskan Gyre region during the 1980­ 89 period was lower than in the 1955-58 PIB was estimated to be 0.47 year-1 by period (Brodeur and Ware 1995). Northward entering a longevity of 9 years (Savinyck movement occurs during summer and and Vlasova 1994) into the empirical pomfret reach the northern part of the Gulf equation of Hoenig (1983), which links of Alaska by September and retreat to the longevity and Z, here assumed equal to PIB south to spawn during winter i.e., they (see Allen 1971). appear to follow the northward movement of

24 Jack mackerel (Trachurus symmetricus) 1974), a mean habitat temperature of 15°C, and a caudal fin aspect ratio of2. Jack mackerel distribution is bounded by the 11 0 C isotherm (Brodeur 1988), and thus ranges further north in summer than in Daggertooth (Anotopterus pharao) winter. Catch data by Brodeur and Ware (1995) shows that jack mackerel were not Welch et al. (1991) present evidence exploited during the 1980-89 period in the showing that Anotopterus pharao, a Alaska Gyre. specialized member of the order Myctophiformes (Fam. Anotopteridae) The diet of jack mackerel, summarized by reaching some 85 cm (TL), attacks adult Brodeur (1988), consists primarily of Pacific salmon (Oncorhynchus spp), and euphausiids, which make up over 50% of may be able to ingest juveniles. stomach content by weight, the rest being contributed by copepods, decapods, They further report that, "as up to 12% of pteropods, cephalopods, and fish. adult sockeye salmon returning to British Columbia bear slash marks, the potential Key parameters are PIB = 0.5 year-I, and significance of A. pharao as a cause of Q/B = 7.0 yea{l (Jarre-Teichmann and mortality for juvenile salmon needs to be Christensen, in press). evaluated".

Unfortunately, no data appear to exist on the Saury (Cololabis saira) abundance ofA. pharao in the North Pacific, or on its diet, and it is thus difficult to follow Saury is the dominant small pelagic fish in up on this suggestion here. the subarctic zone in summer (Brodeur 1988). This species is not well-sampled by the gill nets used on the high-seas, and its Marine Mammals abundance is not known (Brodeur and Ware, 1995). Saury exhibit a preference for water temperatures in the 15-180 C range (Andrew Trites and Kathy Heise) (Kasahara 1961), and is not present in the Alaskan gyre during winter. Thirteen species of marine mammals feed in the Gulf of Alaska gyre during summer and The diet of saury consists primarily of winter. These were grouped into five copepods, which comprise over 50% of the categories: pinnipeds, toothed whales, stomach contents by weight. Euphausiids, baleen whales, beaked whales and killer amphipods, decapods, and fish contribute whales (resident and transient). Estimates of the rest (Brodeur 1988). Also, we have mean body weight (wet, i.e., live weight) for (from HU~hes 1974): Loo = 35 cm (TL), K= males and females of each species were 0.34 year- , and M = PIB =1.6 year'l. obtained from Trites and Pauly (in prep.). Population estimates were obtained from A Q/B value of 7.9 year'l was obtained published sources or educated guesses based using the empirical model of Palomares and on the best available information, such as Pauly (1989), with Woo = 193 g (Hughes Northridge's (1991) global population

25 estimates. Unless otherwise stated, Steller sea lion, feeds in the gyre during individual ration (R, in % of body weight winter. The maximum rate of population l dai ) was estimated for each species and growth for northern fur seals and other sex using: pinnipeds is believed to be about 12% (Small and DeMaster 1995). The PIB ratio 8 R = 0.1·Wi.sO. was therefore set at 6%, half of the maximum. where Wi,s is the mean body weight in kg of species (i) and sex (s), 0.8 is from Eq. 23 in Northern fur seals (Callorhinus ursinus) Innes et al. (1987), and 0.1 is a downward from the Pribilof Islands numbered adjusted value (from 0.123 in Innes et al.), 1,019,192 in 1994 (Small and DeMaster which account for the difference between 1995). Their annual migration extends from ingestion for growth and ingestion for the Bering Sea to the coastal waters of maintenance. California. Much of the population migrates through the Western Gulf of Alaska from Estimates of daily ration ranged from 1.7% April to July (Bigg 1990). Between 10% and of body weight in a 6,100 kg minke whale l 25% of the population feed in the Alaska (107 kg dai ), to 5% of body weight per Gyre on their return to the breeding islands day in a 32 kg harbour porpoise (1.6 kg l in the Bering Sea (10% in April, 25% May, dai ), and are compatible with present 23% June, and 20% July, as calculated from knowledge of the biology of large and small the number offur seals sighted in all areas of marine mammals (Bonner 1989). Dietary the north Pacific during pelagic surveys composition was determined from stomach shown in Figure 8 of Bigg (1990). This and fecal remains reported in published represents 13% of the total population over sources for each species (e.g., Perez 1990). the 6 summer months or 130,000 fur seals Dietary composition for pinnipeds, toothed per month. 5,000 fur seals were assumed to whales, baleen whales and beaked whales be present from October to March, given was set equal to the mean diet of the species that few, if any, appear to be in the gyre within the grouping, weighted by the during winter (Bigg 1990). The ratio of relative population abundance and daily males to females was assumed to be 1:4 in ration estimates for each species. both seasons. Mean body weight of males (30.2 kg) and females (25.3 kg) were taken Detailed summer and winter population and prey composition data for all 13 marine from Trites and Pauly (in prep.). mammal species are summarized in Dietary information was based on stomach Appendix 1, Tables A-C. Additional contents from fur seals shot at sea from 1956 information concerning the assumptions and to 1972 (Perez and Bigg 1981, 1986). The estimates used for each species of marine Gulf of Alaska gyre corresponds to Area 16 mammal in the Alaskan gyre is given below. of the fur seal pelagic survey (Table 19 of Perez and Bigg 1981) where, in the summer, Pinnipeds the animals eat predominately squid (78%), salmon (11 %), rockfish (8%), and pollock Northern fur seals and northern elephant (3%). While no animals have been collected seals are found in the Alaska gyre in both in the gyre during the winter, it is reasonable summer and winter. A third species, the to assume that a few must feed here.

26 Without dietary information it was assumed pollock), based on dietary information that the animals had consumed a generic compiled by Merrick (1994). Mean body diet, taken from Pauly et al. (1995). weight was taken from Trites and Pauly (in prep.). Northern elephant seals (Mirounga angustirostris) make biannual migrations from the breeding beaches in California to Baleen Whales deep waters of the Gulf of Alaska (DeLong Three species of baleen whales are found in et al 1992; LeBoeuf 1994; Stewart and the Alaska gyre during summer months: DeLong 1994, 1995). Males go further north blue, fin and sei whales. Minke and than females and may feed in the Alaska humpback whales are primarily coastal Gyre for 30-50 days of each trip, before species (Leatherwood et al. 1982; Jefferson returning south. It was assumed that 40% of et al. 1993) and were not considered to occur the male population spent up to one month in significant numbers in the gyre. All of the in the gyre in the summer and another month baleen whales show seasonal movements feeding in the area in winter. Given the into southern latitudes in winter months current population estimate of 127,000 (Leatherwood et al. 1982), and are not elephant seals (Stewart et al. 1994) and an present in the gyre in winter. assumed sex ratio of 50%, approximately 4,000 males should be present in each of the The maximum rate of population increase 6 months ofsummer and winter. assumed for baleen whales is 4% (Reilly and Barlow 1986) and production was estimated In the absence of dietary information for to be 2% (half of rmaJJ Much of the elephant seals feeding in the Gulf, it was information on distributions and diet was assumed that they ate 40% small squid, 20% obtained through historical whaling accounts dogfish, 10% rockfish, 10% sablefish, 10% from the coast of Japan, the Gulf of Alaska hake and 10% miscellaneous demersal fish, and the coast of British Columbia (see, e.g., based on dietary composition estimates from Scammon 1874; Townsend 1935; Nemoto elephant seals sampled in California 1959; Nichol and Heise 1992). (Antonellis et al. 1984). Mean body weight was taken from Trites and Pauly (in prep.). Blue whales (Balaenoptera musculus) wintering off the coast of California and Steller sea lions (Eumetopias jubatus) breed Mexico number 1,700 (Small and DeMaster on offshore rocks and islands from 1995), but no estimates are available for California to northern Japan. They generally populations summering further north. The feed within 20 km of shore during summer, Gulf of Alaska is the northern limit of the but venture several hundred km during range of blue whales (Jefferson et al. 1993). winter (Merrick 1995). The western Gulf of An evenly distributed summer population Alaska population numbered approximately size of 1,700 animals (range 1,000-3,000) 15,000 in 1994 (Trites and Larkin 1996) and was assumed. has declined by over 65% since 1980. Winter diet is not precisely known, but was Blue whales consume 40 g of food per kg of assumed to consist of 15% squid, 20% small body weight per day during the summer pelagic fishes (capelin, mackerel, herring), feeding season and increase their body mass and 65% large pelagic fishes (mostly by 50% (Lockyer 1981 b). Dietary

27 information obtained from historical whaling Toothed Whales accounts from Japan and British Columbia show blue whales to feed primarily on Dall's porpoises, killer whales and sperm euphausiids (95%) and occasionally on whales are found in the Alaskan gyre in copepods (5%) (Nemoto 1959; Nichol and summer. The maximum rate of population Heise 1992). increase for all toothed whales is believed to be 4% (Reilly and Barlow 1986) and annual Fin whales (Balaenoptera physalus), like blue production was estimated to be 2% of whales, extend their feeding range into the biomass (halfofrmaJJ Gulf of Alaska during the summer months. Fin whales increase their weight by an Sperm whales (Physeter macrocephalus) are estimated 30% over the summer months, with found in the gyre in summer only, and all a daily ration of about 40g per kg body individuals are mature males. Much of the weight (Lockyer 1981 b). Dietary information information available on the distribution and obtained from historical whaling accounts diet of sperm whales was obtained through indicates that fin whales feed on euphausiids historical whaling accounts from the coast of (75%), copepods (20%), and fish (5%) Japan, the Gulf of Alaska and the coast of (Nemoto 1959, Nemoto and Kawamura 1977, British Columbia (e.g. Townsend 1935, Nichol and Heise 1992). Nichol and Heise 1992). According to Townsend (1935), sperm whales north of Fin whales, once the most abundant baleen 49° N were "stragglers" from the breeding whale in the world's oceans (Evans 1987), groups of sperm whales found further south. were commonly taken by whalers, but are It was assumed that 2,000 sperm whales presently listed as endangered (Small and were present in the Pacific north of 45° N in DeMaster 1995). Current population summer. The average weight of mature estimates range from 17,000 to 20,000 in the sperm whales in the Antarctic was 27.4 t, North Pacific (Evans 1987; Gambell 1985a). and they consumed approximately 3% of their biomass per day (Lockyer 1981 a). Sei whales (Balaenoptera borealis) are These parameters were assumed to apply to primarily an offshore species with an the mature males found in the gyre in estimated population of 14,000 animals in the summer. North Pacific (Gambell 1985b). Like other baleen whales, they move into cooler waters Diet information from the Gulf of Alaska in summer to feed and move into lower was not available, and thus, historical latitudes in winter to breed. It was assumed records of the stomach contents of 501 that they feed at approximately the same whales harvested off the west coast of summer feeding rate as do the other baleen Vancouver Island (Nichol and Heise 1992) 1 whales (4% of body weight·dai , Lockyer were used here. They indicate that sperm 1981 b). Based on historical whaling data, sei whales feed primarily on large squid (80%), whales feed primarily on copepods (80%), but also consume small squid (5%). Fish followed by small squid (5%) euphausiids were also consumed, notably the ragfish (10%) and small pelagic fish (5%) (Nemoto lcosteus aenigmaticus (15%). 1959; Nichol and Heise 1992).

28 Resident killer whales (Orcinus orca) in assumed to occur in the Gulf of Alaska. British Columbia and in Prince William While no diet information is available from Sound, Alaska eat fish (Bigg et al. 1990; stomach contents of whales in this area, it is Heise et al. 1992; Ford et al. 1994). reasonable to assume that it is similar to Approximately 238 resident whales live in transients from other areas. Based on the Gulf of Alaska and the Bering Sea. Diet Barrett-Lennard et al. (1995), summer diet information from stomach contents is not was assumed to comprise 50% toothed available for this area, but there· are many whales (predominantly Dall's porpoises), reports of killer whales raiding commercial 40% baleen whales and 10% pinnipeds. In longline gear in the Gulf of Alaska and in winter, it was assumed that transients spent the Bering Sea. Based on this, and on studies more time foraging in nearshore areas, of killer whales from other areas, resident yielding an estimated population of only 22 killer whales in the Gulf of Alaska gyre are animals, and a diet composition of 60% assumed to eat primarily salmon (80%), as toothed whales and 40% pinnipeds. The sex well as large (10%) and small pelagics ratio for resident and transient killer whales (10%). The winter diet is assumed to contain was assumed to be the same (0.64:0.36 less salmon (60%), and an increased number female : male). Marine mammal prey has of large and small pelagics (20% each). higher caloric value than that of fish (Perez Adjusting for the age structure of the killer 1990), and Barrett-Lennard et al. (1995) whale population and the caloric value of estimate that male and female transients prey, Barrett-Lennard et al. (1995) estimated consume 73 kg ofprey per day. that male and female killer whales consume 84.6 kg and 84.1 kg respectively of food per Dall's porpoises (Phocoenoides dalli) are day, which is slightly higher than would be widely distributed throughout the north predicted from the empirical equation of Pacific. Hobbs and Lerczak (1993, in Small Innes et al. (1987). and De Master 1995), estimated an abundance of 106,000 animals for the Gulf Olesiuk et al. (1990a) estimated a production of Alaska. Applying a correction factor for of 2.92 % year- t for resident killer whales in vessel attraction (0.2, based on Turnock and British Columbia, while Small and Quinn 1991) gives a population abundance DeMaster (1995) used rmax = 4%, a value estimate of 21,200 (range 15,000-30,000) that is considered conservative for most for the Gulf of Alaska. This estimate was cetaceans (Reilly and Barlow 1986). Thus, used for both summer and winter models we assume a PIB ratio of 2% for resident because Dall's porpoises do not appear to killer whales. Barrett-Lennard et al. (1995), show strong seasonal movements (Green et based on data in Bigg et al. (1990), al. 1992). Mean body weight estimates for estimated the sex ratio of female to males to males and females were 63.1 kg. and 61.4 be 0.64:0.36. Mean body weight estimates kg, respectively (Trites and Pauly, in prep.). are 2,587 kg for males and 1,973 kg for females (Trites and Pauly, in prep.). The food consumption of Dall's porpoises was estimated from the equation presented Approximately half of the 88 transient above, adapted from Innes et al. (1987). The (mammal-eating) killer whales that utilize diet of DaB's porpoises from the Gulf of the waters of western Alaska and the Bering Alaska is not known, but Klinowska (1991) Sea (Barrett-Lennard et al. 1995) are lists Pacific mackerel, sardines, saury, and

29 squid as occurring in the diet of animals predominance of squid in the diet, with taken in the western north Pacific. The diet rockfish, skate and sardines occurring less composition retained here is based on the frequently in animals harvested by Soviet stomach contents of28 animals taken offthe whalers. A diet composition of 35% large coast of Japan, analyzed by Wilke et al. squid, 30% small squid, 25% large pelagics (1953), and consisting of mesopelagics and 10% small pelagics was assumed. (70%), small squid (20%) and small pelagics (10%). Stejneger's beaked whales (Mesoplodon stejnegeri) inhabit cold temperate and subarctic waters, rarely ranging below 45° N Beaked Whales in the eastern north Pacific (Jefferson et al. 1993). They are one of the more common The maximum rate of population increase species found stranded in Alaska for all toothed whales is believed to be 4% (Zimmerman 1991), but are rarely seen at (Reilly and Barlow 1986) and the production sea. It was assumed that Stejneger's beaked for beaked whales was set at halfthis value. whales are present year round in the Gulf of Baird's beaked whales (Berardius bairdii) Alaska gyre because there is no seasonality are the largest ofthe beaked whales and may to the strandings. Given no reliable attain lengths of up to 14 m, and ages of up population estimates are available for this to 84 years (Rice 1986; Klinowska 1991). species, density was based on an estimated Barlow et al. (1995) estimate a total of 19 population size of 1,000 averaged across the Baird's beaked whales off the coasts of entire area oftheir distribution, which agrees California, Oregon and Washington. As the with the order of magnitude estimate of species is primarily pelagic, an evenly Northridge (1991). Diet is primarily squid if distributed population of 10,000 was one can trust the two stomach contents assumed across their range (from Jefferson reported in Klinowska (1991 ), and the et al. 1993). However, most animals in the information in Tomilin (1957). north Pacific move south in winter (Tomilin 1957). The sex ratio of adult Baird's beaked Cuvier's beaked whales (Ziphius cavirostris) whales appears to be strongly biased are perhaps the most abundant ofthe beaked towards males; for example, only 3 of 24 whales, with an almost cosmopolitan whales caught off Coal Harbour between distribution. Barlow et al. (1995) estimate 1948 and 1967 were females (Rice 1986). A 886 animals off the coasts of Washington, male to female sex ratio of3:2 was assumed. Oregon and California. A population size of 1,000 animals was assumed with no Rice (1986) found primarily medium-sized seasonal changes in the distribution or squid in the stomachs of animals harvested feeding behavior. Tomilin (1957) reports by California whalers. Klinowska (1991) squid in the stomach of one specimen. cites benthic fish and cephalopods in the which, in the absence of more information, stomachs of animals harvested in the was assumed to be the sole prey of this western Pacific. Tomilin (1957) reports a speCIes.

30 Seabirds in the Alaskan Gyre express an average for the summer months, from April to September (see also Wahl et al. 1989). Body weights were obtained (John Kelson, Yoshihiko Wada and mainly from Palmer (1962), but in some Suzann Speckman) cases, they were estimated from length and related to similar species. Body weights and Sooty and short-tailed shearwaters dominate diet compositions for petrels, jaegars, the avifauna on the outer coast ofthe Gulfof phalaropes and albatrosses were combined, Alaska, both numerically and in terms of assuming that body size and prey species biomass. A variety of prey species are used were similar. A few rare prey species were by seabirds in the Gulf; of these, capelin, omitted from the diets, as were unusual sand lance, and euphausiids are of greatest feeding events, such e.g., as a glaucous­ importance. Trophically, seabirds in the Gulf winged gull eating an ancient murrelet. range from near primary consumers to third­ order carnivores (Sanger 1987), ingesting an Table 8 summarizes our population estimated 1,120,000 t during the 120-day estimates, which lead to an overall biomass 2 summer period (DeGange and Sanger, of 0.0055t·km· , and to Q/B = 101 year'l. 1986). The PIB ratio ofthese birds could not Table 9 presents the diet composition of be estimated using the data at hand, and was seabirds in the Alaska gyre; see Wada (this assumed = 0.1 year' I, based on contributions vol.) for the estimation of consumption in Christensen and Pauly (1993). Values rates.

Table 8 Biological and population statistics of marine birds in the Alaska gyre.

Species Body weight Daily ration Q Population Consumption. (Unit) (kg) (%W) (kg year" ) (N'km""Y (kg'year" km''') Cassin's auklet 0.15 35.2 19.3 0.010 0.19 Common murrelet 0.8 27.4 79.9 0.010 0.80 Arctic terns 0.15 35.2 19.3 0.250 4.82 Tufted puffin 0.6 28.6 62.6 0.010 0.63 Homed puffin 0.55 28.9 58.1 0.010 0.58 Black-footed albatross 3.09 22.3 252.1 0.060 15.12 Jaegars I 26.5 96.6 0.110 10.63 Sooty shearwater 0.79 27.4 79.1 3.500 276.76 Short-tailed shearwater 0.46 29.7 49.9 0.550 27.46 Northern fulmar 0.78 27.5 78.2 1.800 140.80 Glaucous-winged gull 1.16 25.9 109.6 0.400 43.84 Mottled petrel 0.35 31.0 39.6 0.550 21.77 Fork-tailed storm-petrel 0.048 41.7 7.3 1.610 11.77 Leach's storm-petrel 0.048 41.7 7.3 0.550 4.02 Black.-legged kittiwake's 0.2 33.7 24.6 0.060 1.48 Phalaropes 0.03 44.8 4.9 0.005 0.02 Sums -- - 9.49 560.69

31 Table 9 Diet composition ofseabirds in the Alaska gyre.

Food items·J Bird Species Crust. Ceph. Cop. Euph. Dec. S. pel. L. pel. Benth. Cassin's auklet - I.l 76.7 3.5 14 4.7 - - Common murre - 0.1 - 3.2 I 74.5 11.7 - Arctic terns 0.1 - - 95.6 0.1 2.9 - 0.1 Tufted puffin - 7.8 - 11.2 - 80.6 0.6 - Homed puffin 0.2 1.2 - 0.5 - 98 0.1 - Black-footed albatross - 100 ------Jaegars - - - -- 90 10 - Sooty shearwater 0.2 26.6 - 1.5 - 71.8 - 0.1 Short-tailed shearwater 0.1 2 - 72.4 - 23.6 0.1 1.8 Northern fulrnar 0.7 96.3 - 0.2 - 2.2 0.6 0.2 Glaucous-winged gull 0.2 - - 0.4 0.3 95.4 - 1.6 Mottled petrel - 70 - 20 10 --- Fork-tailed storm-petrel - 60.7 - 22.3 3 4.2 1.7 1.3 Leach's storm-petrel - 70 - 20 10 - -- Black-legged kittiwake's 0.6 0.1 - 10.4 0.2 79.9 1.2 2.2 Phalaropes - 10 30 30 30 --- Overall contr. (%) 0.3 44.74 0.021 5.93 0.47 47.79 0.4 0.34 a) Crustaceans, Cephalopods, Copepods, Euphausiaceans, Decapods (crabs), Small pelagics, Large pelagics, Benthos.

Balancing the Alaska gyre model a diet composItion composed of 40 % herbivorous zooplankton, 40 % krill, 5% crustaceans other than krill, 5 % squids (i.e., (Villy Christensen) cannibalism) and 5 % pelagic fishes.

Only few modifications had to be made to Another modification was to account for the the input data in order to balance the Alaska fact that the marine mammals could not gyre model. accommodate the predation pressure exerted by the transient killer whales. This was One of these was the introduction of a corrected by making 75% of killer whale "squid box", suggested by their importance diet be an import, i.e., the transient orcas in the diet of various groups. However, feeding in the Alaska gyre system are given the scarcity of information on squids assumed to take 75% oftheir food outside of biomass and related parameters in the North­ that system. This mayor may not be a good eastern Pacific, this box was based largely assumption - what matters is that, with the on model-generated constraints, i.e., squid data available, this was a possible solution. production was estimated from the Another solutions may be that the number of requirements of squid predators. Other killer whalers was overestimated. inputs were a Q/B value set at 15 year-I, and

32 Table 10 Basic parameters for the Alaska fg're model. (Estimated parameters are shown in brackets). Biomasses are given in t·km- ·year-" while productionlbiomass (PIB) and consumption/biomass (QIB) are both annual rates. EE is the ecotrophic efficiency, expressing the proportion of the production that is lost to exports or predation mortality.

Group Biomass P/B QIB EE Phytoplankton 15.200 175.00 0.0 (0.95) Bacteria 11.000 18.50 25.0 (0.45) Microzooplankton 1.700 63.00 210.0 (0.9:3) Small herbivor. zooplankt. 25.000 27.00 90.0 (0.22) Salps 8.000 9.00 30.0 (0.00) Carnivorous zooplankton (0.579) 3.00 10.0 0.95 Jellies 9.100 3.00 10.0 (0.02) Krill (4.648) 3.00 15.0 0.95 Squids (2.630) 3.00 15.0 0.95 Other crustaceans (1.683) 4.00 13.3 0.95 Pink 0.083 2.45 12.2 (0.35) Sockeye 0.109 0.92 4.6 (0.71 ) Chum 0.051 1.64 8.2 (0.84) Steelhead 0.100 1.00 5.0 (0.80) Mesopelagics 4.500 0.70 3.0 (0.62) Small pelagics 1.894 1.60 7.9 (0.95) Sharks 0.050 0.20 10.0 (0.05) Large fish (0.718) 0.47 4.3 0.95 Pinnipeds 0.072 0.06 16.9 (0.65) Toothed whales 0.097 0.02 5.4 (0.78) Baleen whales 0.256 0.02 14.6 (0.24) Beaked whales 0.003 0.02 7.5 (0.00) Orca (transient) 0.001 0.02 12.1 (0.00) Marine birds 0.006 0.10 101.0 (0.00) Detritus 0.000 - - (0.27)

The ecotrophic efficiencies of the marine compositions are given in Appendix 1, mammals is generally on the high side. This Table E. The flowchart ofthe model is given may indicate that their production (biomass in Figure 1, while Figure 2 presents the and/or P/B) is underestimated, or that the mixed trophic impacts within the Alaska predation is overestimated, as discussed gyre ecosystem. above.

The basic parameters for the model are given in Table 10, while detailed diet

33 --+ F\ow 5 -- e-tor --.. Respiration TIIIIIOrcu ..s- Import 0.0

8ea~I... . Sharks 0.0 _I----- 'ToothecIWh'I~ ~ ~ Steelheacl J .0.5:- _ 0.0 t1 r-JPimipecb : I.-.-J ~""'''I -~ 4 ;.[]- ,[r-l' £[..- . - -i[g o~ 2.a •.v T 0.9 I::';:J 0.3

-e e> III ~ " .J Squids t-- sr1 ,..----, II !l.r B.leenWh.les w ..f1 P Mnopel....-ISmPelaBict ~ ,g 3 31.9 2.0 11.6 12.1 3.7 at; Jellies b s:. a. T 4.6 e -r'--.-::I.-_---. to- IT II lt- '-1.5 Salps ...-..L..-~'---= /I II I r-'-....L--"IL-. ~icrozoopl.h ..-IOIherClUs~ r )11 III f'4' "1 I 15 9 'I Fil CamZoopl I 235.5 .. Bacteria 253.9 . .. SmHetbZoopl I-- 4.1 i Krill 2 183.2 2132.6 56.5 L---.,-...J T - I I -1_

Figure 1 Flow chart oftrophic interactions in the Alaskan gyre system. All flows are expressed in t.km-z.year-1 and all biomasses in t·km-z; minor flows are omitted, as are all backflows to the detritus. Note that total area ofsystem is not defined, i.e., the flow chart may be seen either as representing only its core, or include much ofits periphery. IMPACTED GROUI'

8 i!L '~ -i II II -a i!L ~ 1! ~ 1 ~ ~ iii Ii. .g ~ ~ 1 ~ "8 0 ~ ~ Ju u i 11 a :I: l!. ~ II 1l ..w e ~ ~ -5 § ..w 'S ..... U u go .., Iu .<: e = " ] "e .., :§ II j ·1 ~ ~ ~ ~ ~ ~ ~ j ~ ~ ~ ... a:l" '" '" 5 U '" ::E '" '" ... a:l ::E

Phytoplankton IiltWIWWiiU Baeleria ~ """""" Mierozoopl. i'i.W.liiIW!llIIID SmHcrbZoopl IlllilIill -lFl!!B1llIIlBII1IIifiIilI IIImlflS ......

Salps ~"1a ~ ~ -- CDmZoopl Jellies ... ~-- --- ImlllmJ ::> Krill 0 cr: Squid. 0 -r-"1liIIlllllll -- --_ 0 Olh.rCrusl :z I- Pink U ...< Scxkcylo: -::E Chum Steclhead I;J Mesopclagies VI lIIlllUI Sml'clagies ...... Sharks Ilm!IlIiO WlllllS Lar@eFish - l'iMipcds -_-- tr=iI ToolhedWhales .

naleenWhales

neakcdWhales

Tr.nsOrc.~ MarineBirds •• Octrilus

Figure 2 Mixed trophic impacts in the Alaska gyre ecosystem, generated by Ecopath, and illustrating how an infinitesimal increase in one of the groups would affect the other groups. (The scale is relative, but may be used for comparisons among groups). Note indirect effects, indicative of trophic cascades. The model was easily balanced at the lower an explanation for the continuous absence of trophic levels, and the balance thus obtained phytoplankton blooms in the offshore seems very plausible. It is interesting though regions of the subarctic Pacific (Miller that the EE ofphytoplankton is very close to 1993). The absence of blooms is well in line 1, indicating a full utilization of the with the present finding that the zooplankton phytoplankton within the system. On the is very productive, and capable of other hand, the EE of small herbivorous constraining the phytoplankton. zooplankton is only 0.22, so that less than a quarter of the small zooplankton production The above considerations are only meant to is utilized. It is not clear if this is an artefact illustrate the kind of results that can be [shifting a larger proportion of the obtained from Ecopath modelling By zooplankton feeding to large zooplankton placing the data in a rigid network, (chaetognaths) would have increased the information is obtained. Some parameter biomass of these, which in tum would have combinations are simply not possible, and lead to a demand for, and hence higher the present approach is indeed good for utilization of the small zooplankton]. identifying inconsistencies. Much more However, artificially shifting a major part of could be done using tools for analysis built zooplankton feeding toward the larger into the software (see, e.g., Figure 2).; we zooplankton is not enough to strongly abstain, however, from presenting more increase the utilization of small herbivorous here, until a winter model has been zooplankton. Their concentration and constructed for the area. Indeed, we expect production is simply too high. that much would be gained by explicitly considering seasonal oscillations of mass­ Interestingly, a major purpose ofthe SUPER balance models, e.g., in the framework project mentioned above earlier was to seek presented by Walters (this vol.).

36 SOUTHERN B.C. SHELF Euphausiids, Chaetognaths and Herbivores MODEL

(Jenny Purcell) The area covered Euphausiids (D. Pauly) A first estimate of euphausiid biomass can The southern coast of British Columbia as be obtained by averaging the entries of Table 2 in Fulton et al. (1982), and this defined here ranges from nearshore waters 2 (about 10 m) to the shelf slope (300m), from amounts to 3.8 ± 7.3 g wet weight·m- . the southern tip of the Queen Charlotte Another rough estimate is 0.65 g dry weight·m-2 (Mackas et al. 1992), corres­ Islands in the North, along the Pacific Ocean o2 coast of Vancouver Island to the Juan de ponding to about 2.8 g wet weight.m Fuca Strait in the South, and thus excludes (Sambilay 1993). the Hecate, Charlotte, Johnstone, and Georgia Straits. Chaetognaths The system thus defined covers 2 A very approximate biomass for approximately 30,000 km , and is, in fact, so chaetognaths of 0.3 g dry weight m-2 was open to adjacent water bodies that it cannot obtained from Mackas et al. (1992), and this legitimately be considered a distinct translates to 3 g wet weight· m-2 if we ecosystem. However, recurrent groups of assume dry weight = 10% wet weight. The interacting species of fish and invertebrates diet consists almost exclusively of do occur in the area thus defined, and zooplankton (mostly copepods). various biomass and flux estimates have been published by authors who (at least implicitly) must have assumed some degree Salps of integration among these groups. Thus, for pragmatic reasons, we shall treat the above­ Biomass is 0.5 ± 0.5 g.m-2 dry weight (from defined area as if it identified the boundaries Mackas et al. 1992); if we assume dry of a mass-balanced ecosystem, but return weight equal to 10% of wet weight, then later to the open nature of the "ecosystem" biomass is about 5 g mo2 (wet weight). defined here. The diet probably consists predominantly of Primary production in the area has been phytoplankton, with a minor micro­ estimated as 345 gCm-2 year- 1 by Robinson zooplankton contribution. et al (1993), who also reviewed the oceanography ofthe Southern B.C. shelf.

37 major groups present (c.f. with Kozloff Copepods 1987). Thus, species composition in anyone area depends on the substrate as well as on Biomass was estimated as 1.825 ± 0.9 g dw -2 the oxygen condition. Polychaetes are m by Mackas et al. (1992, Figure 9). If we dominant, bivalves are less widespread, but assume dw = 10.9% wet weight (Sambil~l still common. Ostracods, harpacticoids, 1993), then biomass = 16.7 g wet weight m . cumaceans, tanaids and amphipods dominate the small benthic crustaceans. Brittle stars, a The diet (in the Gulf of Alaska) is a mix of heart urchin and a holothurian were also phyto- and microzooplankton, viz. 91.5% present. phytoplankton and 8.5% microzooplankton. In the absence of local estimates, Brey's (1995) database on macrobenthic Invertebrate Benthos productivity was used to obtain rough estimates ofP/B and related statistics for the (Astrid Jarre-Teichmann and major benthic groups. Sea stars, sea urchins Silvie Guenette) and sea cucumbers, benthic cnidarians and sponges were not considered, because their A survey in the mid-1980s (Brinkhurst role as prey items to fish appears to be very 1991) revealed that the benthic fauna of the limited. Table 11 lists the most abundant southern B.c. Shelf is complex, with all benthic groups and their P/B estimates. Table 11 Benthic groups and their PIB ratios, southern shelf ofB.C.

Functional Abundant species on the southern B.C. Species for which PIB estimates PIB group shelf a exist in a similar temperature adopted range (point estimate, or range; (year-I) year-I) b. Polychaetes Mediomastus ambiseta, Galathowenia Euzonus mucona (1.8), 2.0 oculata, Prionospio steenstrupi, Pectinaria californiensis (3.2-5.4), Aricidia ramosa, A. lopezi, Levinsenia Paraprionospio sp. (1.2-2.0), gracilis, Cosssura soyeri, C. sp., Aricidia spp. (1.3-2.6), Spiophanes Spiophanes berkleyorum, Glycera bombyx (1.5-4.7), Glycera alba (l.0), capitata, Sphaerosy//is brandhosti, Chaeozone setosa (1.3) Tharyx seconds, Nototmastus /ineatus, Chaetozone spp. Bivalves Axinopsida serricata, Adontorhina cyc/ia, Yoldia thraciaeformis (0.8) 0.7 Yoldia scissurata, Y. thraciaeformis, Y. notabilis (0.4-0.8) Macoma e/imata, M. car/ottensis, Macoma balthica (0.8) Huxleya minuta, A. cyc/ica Tegulafunebra/is (0.4) Small crustaceans Cumacea: Eudore/la pacifica, Cumacea: 2.5 Leucon nasica, Lamphrops serrata Diasty/is rathkei (2.0-2.7) Tanaidacea: Cryptocope spp. Amphipoda: Amphipoda: Harpiniopsis spp., Harpinia propinqua (3.8) Heterophoxus oculatus, Amphelisca spp. Ampe/isca spp. (0.9-5.2) Brittle stars Amphioplus macraspis, A. strongyloplax, Ophiura spp. (0.3-1.3) 0.6 0. sarsi "from Binkhurst (1991) bfrom Brey (1995) and references therein

38 Table 12 Suggested production:biomass As Q/B estimates were available for neither ratio (PIB; year-\ gross conversion of the benthic groups, they were estimated efficiency (GE), and ecotrophic efficiency (Tablell) based on a range of mean gross (EE) for major benthic groups of the Southern B.C. shelf. efficiencies of 9% for herbivores/ detritivores, and 30% for carnivores (Brey Group P/B GE EE 1995). Polychaetes 3.0 0.09 0.95 Bivalves 0.5 0.09 0.80 With the exception of shrimps, no benthic Amphipods 2.4 0.20 0.95 biomass figures were available. As a guide Brittle Stars 1.8 0.20 0.80 to balancing the model, we suggest: B Sea urchins 0.4 0.09 0.60 (polychaetes) > B (medusae) > B Sea stars 0.4 0.09 0.60 Shrimps 0.7 0.15 n.a. (amphipods). Table 12 presents our Crabs 1.8 0.25 0.95 estimated Production: Biomass ratio (P/B), Gross conversion Efficiency (GE), and Benthic shrimps (Pandalus jordani) are Ecotrophic Efficiency (EE) for major commercially exploited off the west coast of benthic groupings. Canada (Boutillier 1991). Their biomass is 2 estimated at 0.3 g m- , and their production The assumed diet composition of benthic 2 1 at 0.7 g·m- ·year- • Their diet composition, invertebrates in the Southern B.c. shelf based on Dahlstrom (1970) was assumed at region is summarized in Table 13. 30% polychaetes, 25% amphipods, and 35% detritus. Carnivorous Jellies Table 13 Diet composition used for the benthic groups in the model ofthe southern B.C. shelf (from Brey, 1995). (Mary Arai)

Group in model Diet composition Summer dry weights for ctenophores, (% weie:ht or volume) Polychaetes 100 detritus hydromedusae and siphonophores on inner Bivalves 100 detritus shelf banks and off southwestern Vancouver Small benth. crust. 90 detritus, 10 zooplankton Island were estimated as 2.6 g dry 2 Benthic shrimps 35 detritus, 35 amphipods, weight.m- (Mackas 1992). Dry weight was 30 Dolvchates. assumed to be 4.2% of wet weight (Larson Benthic crabs 60 detritus, 20 bivalves, 15 benthic shrimps, 3 benthic 1986), leading to a biomass for the crabs, 2 other demersals carnivorous jelly group of 6.19 g wet Sea stars 30 detritus, 40 molluscs, -2 30Dolvchaetes weight·m. This is probably an Brittle stars 80 detritus, 18 small benth. underestimate, as scyphozoa, which are not crust., 2 Dolvch. well sampled by Bongo and other zooplankton nets, are not included. [Dense Dungeness crabs (Cancer magister) are also swarms of scyphozoan occur in late summer exploited off British Columbia. Their total in the neritic zone off Oregon and southern mortality was estimated at 0.7 year-I, and Washington, and contain at least 80% as their diet composition as 17% shrimps, 20% much carbon as the corresponding copepod bivalves, 3% cannibalism, 2% fish and 58% concentrations]. detritus (based on Stevens 1982).

39 PIB estimates range from 7 to 30 year-I Adult krill biomass was estimated at 3.8 2 2 (Larson 1986), while the minimum estimate, t·nm- , corresponding to 1.1 g m- during an based on a generation time, is 6 yea{1 + echoacoustic survey in Jervis Inlet on the mortality rate. For these and other gelatinous British Columbia coast (Romaine et al. animals, the "degrowth" known to occur 1995). This estimate may be used as the under low food conditions, is not considered initial biomass value for the model, although here. it is not clear how representative it is for the whole ofthe area considered here. An approximate diet composItIOn for the group as a whole was derived from data in Hirota (1974), and Purcell (1990), viz Small pelagics • Small herbivorous zooplankton 62% (Rik Buckworth) • Eggs oflarge zooplankton 28%

• Carnivorous zooplankton 5% Herring • Fish eggs 3% Pacific herring Clupea harengus pallasi is • Other items 2% the subject of important roe, spawn on kelp and bait fisheries in British Columbia. It has In winter, hydromedusae are very scarce on been overfished in the past, and is also the B.C. shelf, while ctenophores are subject to considerable natural variation in virtually absent; Scyphozoa would also stock size and recruitment. disappear in early fall (see also Mills 1981). The occurrence of Valella valella along the The Summer feeding grounds are on the B.c. coast is briefly discussed in Arai (this continental shelf (i.e., our reference area), vol.). but in early Fall, herring migrate inshore, where they are subject to intense fishing and predation (Hay and Fulton 1983). The Krill survivors then return offshore, where they stay from February to May. The adults are (Astrid Jarre-Teichmann) pelagic planktivores, feeding mainly on euphausiids, copepods and decapod larvae. Thysanoessa spinifera and Euphausia The biomasses used here refer to 1985-1989, pacifica are the dominant krill species of the were taken from Stocker (1993, Figure 10), southern British Columbian shelf. Their and were estimated from historical catches population biology has been studied on La for the inshore spawning areas. (DFO stock Perouse Bank off Vancouver Island assessment regions Northern West Coast (Tanasichuk 1995). T spinifera is more pro­ Vancouver Island and Southern West Coast ductive than E. pacifica, with estimated PIB Vancouver Island). Natural mortality was ratios ranging from 1.6 to 3.7 yea{l. The estimated as M=0.6 year-I. The stock are energy balance of the krill box, described managed under a constant (20%) harvest rate above for the Alaska gyre model was used in policr The Production:Biomass ratio of 2.2 the absence ofmore detailed information. year- was obtained from Z= F + M = P/B.

40 Diet items mentioned by Stocker (1993) are "abundant through B.c. in a wide range of euphausiids, copepods and decapods. habitats". No information was found that Curiously, for such an important commercial described the species distribution by depth. species, no information was found on the The diet consists mainly of copepods (Hart relative composition ofthe diet. 1973).

Sardines Pacific Cod and Sablefish Pacific sardine, Sardinops sagax was historically the subject of a fishery (around (Patricia Livingston) 20,000 t year- I in the late 1920s) in British Columbia (Hart 1973), but there has not been a fishery since the 1940s. Schweigert Pacific cod (Gadus macrocephalus) (1988) suggests· that S. sagax was "a Pacific cod is widely distributed on both transient visitor" to Canadian waters, from sides of the Pacific Ocean (see Tokranov the northern Californian stock, which and Vinnikov 1991), though it is near the appears to be rebuilding. Villavicencio and southern limit of its distribution off the west Muck (1983) provide some information on coast ofthe British Columbia shelf (Ketchen ration and growth efficiency in this species. 1961). The main concentrations of cod in British Columbia waters south of 52° N are Anchovies found in Queen Charlotte Sound and on La Perouse Bank. Spawning occurs in winter Northern anchovy Engraulis mordax is not (January to March) and the main feeding sufficiently abundant to provide the basis of areas are occupied during April through a fishery in B.C., though it was abundant September. Preferred temperatures in this during the 1940s (Hart 1973). A literature area range from 6.4° C to 7.9° C (Westrheim search produced no information on this and Tagart 1984). Large fluctuations in species in Canadian waters. recruitment produce a great variability in annual landings (Tyler and Foucher 1990). Sandlance In the 1980s, the fishery was closed during winter and the bulk of the landings occurred Although there are large sandlance fisheries during April-September (Tyler and Foucher in the north Atlantic and west Pacific, 1990). Pacific sandlance (Ammodytes hexapterus) is not fished in the northeast Pacific and no Stock assessment of cod indicates an biomass estimates for the species were average catch of 1,695 t from 1985 to 1992 found. The species is an important food item in DFO areas 3C-3D, i.e., west of for a suite of species, for example rhinoceros Vancouver Island (Stocker 1994). Mean auklet (Cerorhinca monocerata; Bertram total mortality during this same period was and Kaiser 1993), Pacific halibut Z = 1.2 year-I. With M = 0.65 year-I, F is (Hippoglossus stenolepsis; Best and St­ about 0.55 year-I . Average biomass for 1993 Pierre 1986, and see Venier, this vol.), was 8,500 1. Ration ranges from 0.9-1.3% salmon, lingcod and marine mammals (Hart body weight daily (Paul et al. 1990), and 1973). Hart (1973) records the species as Q/B may thus range from 3.3 to 4.7 year-I.

41 Diet analysis from Hecate Strait (Tyler and about 17% of the total biomass in the Crawford, 1991) shows that cod consume shallow depth zone ofthe survey, or 2,865 t. mostly herring (60-75% by weight), Many of the juveniles that inhabit inshore demersal fish (10-40%), and sandlance (2­ waters of British Columbia migrate north to 15%). Diet samples of cod taken from 1950 Alaskan waters, and are recruited to the U.S. to 1980 by Westrheim and Harling (1983) fishery there (McFarlane and Beamish show highest occurrences of sandlance in 1983a). Estimates of juvenile biomass from the stomachs of cod from Queen Charlotte the large 1977 year class range from 30,000 Sound and west of Vancouver Island to 60,000 t in inshore waters, including the (frequency of occurrence: 59-83%). Herring shelf waters of Queen Charlotte Sound and (39-43%), euphausiids and shrimp (15­ the La Perouse Bank area (McFarlane and 19%), and sablefish (1-5%) were also Beamish 1983). consumed by cod in those two areas. Off Vancouver Island, there are indications that Estimated natural mortality rate for adults is herring might be the more important prey of 0.08 year-! (Stocker 1994) and the average cod during summer and fall, while sandlance natural mortality rate for juveniles between might be more important during winter and age 0 and age 4 is about 0.6 year-! spring (Westrheim and Harling 1983). (McFarlane and Beamish, 1983a). A Q/B value for adults of 3.73 yea{l was Sablejish (Anoplopoma fimbria) estimated using the equation of Pauly et al. (1993), assuming an asymptotic weight of Exploitable concentrations of sablefish 4,392 g (Stocker 1994) and an annual habitat normally occur between 150-1,000 m depth temperature of 9° C. Q/B for juveniles was (Low 1976). These are long-lived fish (up to estimated to be 6.6 yea{l, based on the 35 years), with rapid growth until maturity. ration estimates for juveniles presented by Adults, aged 3 and greater, are found in McFarlane and Beamish (1983a). slope waters of the continental shelf. Juveniles up to 2 years ofage inhabit surface Juvenile sablefish diet by weight in Queen and inshore waters down to a depth of 150 Charlotte Sound and Hecate Strait consisted m. Most catch in Canadian waters in recent of 41 % herring, 25% euphausiids, 20% years was at depths of 450 m to 990 m. unidentified fish and a total of 14% of crab, Therefore, for the purposes of the model, shrimp, jellyfish, and squid (McFarlane and catches should be set equal to zero. Biomass, Beamish, 1983a). Juveniles are consumed by as estimated during surveys conducted in halibut, adult sablefish, cod, lingcod, and 1992 in the shallowest depth stratum (270­ spiny dogfish. Adult sablefish diet consists 448 m) off the West coast and Queen mainly of fish (50% by weight) and squid Charlotte Sound areas were 7,167 t and (39%). Herring was the main fish species 9,885 t, respectively (Stocker 1994). consumed in summer. Rockfish and myctophids were also identified in the Assuming sablefish are evenly distributed stomach contents of adults from Canadian within this depth zone, and given that the waters. Jellyfish, shrimp, crab, euphausiids, model below only includes depths to 300 m, and amphipods make up the remainder of adult sablefish biomass in the model area is the diet (McFarlane and Beamish 1983b).

42 Pacific Halibut Halibut congregate to spawn and two major spawning areas are located on the British Columbian shelf: one at the northwestern tip (Judson Venier) of the Queen Charlotte Islands and one at their southern tip (Trumble et al. 1993). The Introduction length at 50% maturity for females is approximately 120 cm (Trumble et al. Pacific halibut is an extremely important 1993). resource of the Northeast Pacific - so important, indeed that the fishery is regulat­ Fertilized eggs become part of the plankton ed by its own International Halibut Fisheries where they hatch and tum into juveniles. Commission (IHFC), created in 1923. The planktonic phase lasts for about 6-7 months (Trumble et al. 1993). Juveniles The species is found over the continental settle in shallow water in late spring and shelf of North America from Santa Barbara, early summer (St-Pierre 1989). They California to Nome, Alaska (IPHC 1987). undertake long migrations to the east and Peak abundance occurs in the central Gulf of south, and at 2-3 years, move offshore Alaska (Deriso et al. 1985) from the Alaska (Trumble et al. 1993). These are believed to Peninsula in the north to Cape Fairweather be countermigrations, which may balance in the south (Trumble et al. 1993). the dispersal of eggs and larvae (Trumble et Abundance declines as one moves away al. 1993). Net migratory movements of from this area and becomes limited south of adults, as a result, are small (Trumble et al. southern British Columbia (Trumble et al. 1993). 1993). The species is demersal and is often associated with bottom features such as Pacific halibut can attain lengths ofover 3 m banks and channels, with water temperatures and weigh up to 300 kg (Trumble et al. close to 5° C (Trumble et al. 1993). In 1993). Sexual dimorphism occurs, with the summer, halibut are found in waters greater females growing larger than the males than 90m in depth and in winter they move (Trumble et al. 1993). The oldest halibut on into the deeper waters (300-600m) of the record are a 42 year old female and a 27 year upper continental slope (Trumble et al. old male (IPHC 1987). Fish in the Pacific 1993). commercial setline fishery average 10-14 years of age (Trumble et al. 1993). Growth Seasonal migrations do occur for of both males and females is continuous commercially sized fish (~81.3 cm) throughout their lives and shows little (Trumble et al. 1993). Fish move from decrease at older ages (Trumble et al. 1993). summer feeding grounds inshore to winter Both sexes grow at roughly the same rate up spawning grounds offshore (Trumble et al. to about 5-9 years after which females tend 1993). These movements are usually within to grow faster (Trumble et al. 1993). the IPHC's statistical areas (Deriso and Quinn 1983; Quinn et al. 1985). Predators andpreys Reproduction and growth Pacific halibut are rarely preyed upon by Spawning is an annual event and occurs in other fish and cannibalism is very low late fall and winter (Trumble et al. 1993). (~7%), even in nursery areas (Best and St-

43 Pierre 1986). Some marine mammals, such 1888 in the Washington-Oregon area and as Steller sea lion (Eumetopias jubata), are expanded northward in the early 1900s known to take halibut offofsetline gear, but (Trumble et al. 1993). Commercial catches predation in more natural settings is now average more than 30,000 t per year probably low (Best and St-Pierre 1986). and are highest in the Gulf of Alaska, where the greatest concentration of halibut occurs Juvenile halibut tend to consume a majority (Trumble et al. 1993). In addition to the of invertebrate prey but undergo a transition commercial harvest, halibut became a to piscivory as they grow (Trumble et al. popular sport fish in the 1980s (Trumble et 1993). When they reach about 70 cm al. 1993). The recreational catch constituted standard length (SL), their diet consists about 5% ofthe total directed halibut harvest almost exclusively of fish (Trumble et al. or 3600 t in 1990 (Trumble et al. 1990). 1993). The diversity of prey taken is large Again, the majority of the recreational catch but Pacific sand lance (Ammodytes (75%) occurs in the GulfofAlaska. hexapterus) and Pacific herring (Clupea pallasii) made up 60% and 23% of the diet The International Pacific Halibut in terms of biomass in British Columbia Commission (IPHC) was formed as a result waters, respectively (Best and St-Pierre of petitions from fishers and processors, 1986). whose resource was declining, to the governments of Canada and the US Q/B was estimated at 1.73 year- 1 using an (Trumble et al. 1993). Currently, halibut can empirical equation in Christensen and Pauly only be kept if caught on hook and line gear (1992b, p. 14), with Woo set at 300 kg during open seasons. Set nets were outlawed (Trumble et al. 1993). in 1938 for halibut fishing (Bell 1981) since Best and St-Pierre (1986) estimated the diet they were seen as targeting large spawners composition (in numbers) of Pacific halibut, (Trumble et al. 1993). Trawling for halibut based on analysis ofstomach contents for 70 was banned in 1944 (Bell 1981), on the fish from Hecate Strait and Dixon Entrance. basis that it tended to catch halibut under the Here, data on 40 of their fish from Hecate legal size, and to decrease yield per recruit Strait were used, along with estimates of the (Trumble et al. 1993). weight of prey items, also from Best and St­ Trumble et al (1993) estimated an Pierre (1986), to re-express the diet exploitable biomass estimate of 141,000 t composition in terms of weights. This led to for the entire distribution area of Pacific the following approximate values: sandlance halibut, from Santa Barbara, California, to 50%, herring 20%, miscellaneous demersals Nome. Alaska. Thus, total biomass for the 18 %, crabs 10%, and octopus 2%. coast of southern B.C. was obtained by assuming proportionality between biomass Catches and mortality rates and catch, of which 10% is taken off southern B.C. This, given our reference area 2 Pacific halibut was an important target of 30,000 km , corresponds to a 2 species even before Europeans arrived on biomass/unit area of0.473 t·km- • the west coast, as the First Nations used 2 I halibut for both subsistence and barter (Bell A catch of 0.113 t·km- .year- was derived 1981). Commercial fishing began around from the B.C. landing data presented in

44 Table 4 of the 1990 IPHC Annual Report, The remaining 11 % is composed of many following division by 3 to account for the benthic organisms, none comprising more fact that our reference area represents about than 2% of the diet (Jones and Geen 1977). 1/3 of the B.C. shelf. These landing data, Dogfish appear to consume twice as much expressed in pounds for dressed head-off food in summer as in winter, with annual halibut, were converted to round weight by average consumption of 5 times body weight multiplication with 1.33 for small, and 2.5 times body weight for large animals (Jones and Geen 1977). Thus A natural mortality (M) value of 0.2 year-I average (annualized) summer and winter was given by Trumble et al. (1993), which food consumption rates for this group would seems high, given that these are large, long­ be 5.0 and 2.5 times body weight, lived fish (see Pauly 1980). Fishing respectively. mortality (F) was estimated at 0.24 yea{l, using F = catch in weight I biomass. The definitions Hake

F+M=Z=P/B (Tony Pitcher) were then used to estimate PIB = 0.44 year- t (Allen 1971). Introduction

One of 12 important species of hake world­ Spiny Dogfish wide, Merluccius productus, the Pacific hake or "whiting", is the most abundant (Jeffrey Polovina) commercial fish species along the Pacific coast of North America between 25 0 N and Spiny dogfish (Squalus acanthias) is a 51 0 N (Alheit and Pitcher 1993) Living coastal species ranging from Alaska to principally offshore in 200-500 m depths California. The coastal biomass is estimated over the continental shelf and its edge, the at 280,000 t (Saunders 1988), with about adult hake stock extends from California 150,000 - 200,000 t residing along the northwards through British Columbia in the Canadian coast (Stocker 1994). Using an summer, sometimes as far as the Alaskan estimate of the area of the Canadian shelf of border. It is considered that most hake in 100,000 km-2 gives a density estimate of 1.5 Canada migrate south to US waters in the 2 - 2 t·km- . The PIB ratio is estimated at 0.1 winter (Stauffer 1985; Smith et al. 1992). t year- , and consist mainly of natural Spawning and juvenile hake are generally mortality (Wood et al. 1979). confined to waters south of latitude 44 0 N, although in warm years there are reports of Stomach contents from a large age and eggs in Canadian waters (Beamish and seasonally averaged sample of dogfish off MacFarlane 1985). The extent of the the British Columbia coast indicate a wide northwards extension of 3+ hake depends on range of prey consisting of herring (22%), the warm water associated with ENSO, so euphausiids (14%), flatfish (6%), smelt the proportion of the stock biomass found in (6%), octopus (3%), combjelly (2%), Canadian waters varies considerably from sandlance (2%), and squid (2%). year to year. These coastal Pacific hake

45 exhibit hundred-fold differences In year (Beamish 1979; Venier and Kelson, this class strength which appear to be vol.). uncorrelated with spawning biomass. Longevity, of up to 15+ years buffers the The diet of the offshore hake consists stock against large fluctuations in total principally of krill (= Euphausiidae), but biomass, but hake scales preserved in although fish and shrimp usually make up a anaerobic sediments off California reveal a smaller percentage of the diet, hake are history of large variations in biomass over opportunistic ambush predators (Pitcher and hundreds ofyears (Soutar and Isaacs 1974). Alheit 1995) and at certain seasons and locations, larger Pacific hake may consume In the USA some 250,000 tonnes ofhake are substantial quantities of fish, such as herring landed annually, while the Canada takes or smelts. Pacific hake generally feed about 100,000 t, with catches following a nocturnally, and, surprisingly, compared to generally increasing trend over the past 10 other species of hake world-wide (Alheit years. The fishery, which is considered fully and Pitcher 1995), where hake diet consists exploited, has recently been reviewed by of up to 40% hake, there are no reports of Methot and Dorn (1995); earlier work is cannibalism by M productus. Methot and summarized in Dark (1975). Advances in Dorn (1995) consider that this is because technology now allow production of frozen larger hake migrate out ofnursery areas. fillets, blocks and surimi from a fish that traditionally spoilt rapidly due to proteolytic On account of the importance of the fishery, myxosporidian parasites in the flesh. The the ecology of Pacific hake is relatively well principal issues in assessing and managing documented, with at least 33 papers having the fishery are the influence of ENSO on the been published since 1980. This brief northwards extension ofthe stock; allocation account is confined to an evaluation of the between US and Canada; allocation between literature in order to derive estimates of onshore and offshore processing; and a by­ stock biomass (B, for 3+ adults), total catch of salmon, rockfish and herring, that mortality rate (Z = PIB ratio; Allen 1971), composes a small proportion of the hake and diet, in summer and winter, for input to catch but may be large in absolute amount. an Ecopath model ofthe southern B.C. shelf.

There are three stocks of Pacific hake in Biomass estimation Canadian waters. In addition to the large offshore hake stock that migrates from the For stock assessment purposes, Pacific hake south to summer in Canadian waters, there is biomass is estimated using two a small resident offshore stock that spawns complementary techniques: stock surveys in Canada and whose eggs and juveniles are conducted every three years (these have have been found by surveys in the Barkley been more frequent in some areas), and a Sound area (Beamish and MacFarlane sophisticated catch-at-age model analysis is 1985). Isolated from the offshore hakes, an run every year. inshore, relatively under-exploited stock of hake inhabits the Straight of Georgia and First, every three years, acoustic surveys are Puget Sound, is distinguished by its otolith conducted along the shelf in both countries, structure and lack of flesh parasites supported with suitable trawl verification of

46 species identities. Not surprisingly for an curves are employed for the two countries. ambush predator, most hake aggregations The model is used to project hake biomass contain few other species. Acoustic surveys forward under a range of management have been conducted every year in the options. Risk is evaluated using Monte Canadian zone since 1990. The three Carlo simulation. principal uncertainties in these acoustic estimates are the geographical extent of the Methot and Dom (1995, Figure 7) noted an surveys, which can miss hake offshore and encouragingly close agreement between to the north of Vancouver Island, the estimates of Pacific hake biomass from the identification of sparse hake aggregations in catch-at-age analysis and survey results from the weak scattering layer, and the low target 1983 - 1989. However, later analyses appear strengths ofjuvenile hake. to have eroded this confidence that stock was accurately assessed, and, for later years, Secondly, a rather complex catch at age considerable inconsistencies and analysis, tuned by survey biomass and age uncertainties have become apparent. Much composition, is performed. Ages are based of the information from Canada is in on otolith readings (Beamish 1979), and PSARC (Pacific Stock Assessment Review subsamples raised to a catch-at-age matrix Council) reports, which are labeled as using an age-length key and total catch data unciteable. My evaluation here is a digest in the usual way (Kimura 1989). The current and evaluation that inevitably considers this algorithms are based on a stock synthesis unciteable material. model (Methot 1990), in which age specific mortality is assumed separable into age­ It is clear that acoustic surveys in the 1990s specific selectivity and fishing mortality for have demonstrated hake much further each year (an approach derived from the offshore on the shelf break than had been "separable VPA" of Pope and Shepherd previously realized, and in warm years the 1982). The analysis is said to be "integrated" distribution extends further north than because of the large number of factors taken expected. Also, the introduction of new in consideration when tuning. Tuning is acoustic gear in the US has compounded carried out using "emphasis factors" that calibration uncertainties. determine how closely the expected values from the model approach the observed catch As a consequence, biomass estimates in figures, goodness-of-fit being measured with former years have been retrospectively a log-likelihood function. The fitted revised upwards to compensate for the areas components ofthe model, each of which has previously not covered by surveys: for an attached emphasis factor, include four example the 1986 biomass appears at 2.25 equally weighted estimates of age million tonnes in Methot and Dorn (1995), composition, eight acoustic survey measures but has been revised (by M.W. Dorn) from NMFS in the US, two acoustic upwards to 3.95 million tonnes in an measures from DFO in Canada (given very PSARC document dated 1995. This revision low weighting), and four measures from US makes a large difference to estimates of and Canadian trawl surveys, also given low fishing mortality. Moreover, in 1995 emphasis. An age-specific migration model PSARC documents, we find different figures partitions the stock into Canadian and US to those issued in earlier years. A further components and separate gear selectivity complication is that in most PSARC reports,

47 Table 14 Summer Pacific hake biomass hake assessments. Some of the methods 3 (t.l0 ) and related statistics in Canadian employed are so complex and incompletely coastal waters from the US border to Queen documented that only a complete Charlotte Sound. reconstruction of the modelling exercise O Year Biomass·) CatchUI Fe) M ) could enable an outside reviewer to evaluate 1983 450 41824 0.10 0.22 it confidently. It may be noted that 1986 500 55653 0.12 0.22 alternative models of the Pacific hake stock 1989 225 99532 0.58 0.23 that include environmental drivers have been 1990 316 76680 0.28 0.24 published (e.g. Swartzman et al. 1987), b'ut 1991 402 104522 0.30 0.24 1992 1101 86370 0.08 0.25 appear not to have been considered in 1993 750 58783 0.08 0.25 assessment. Nor have simpler models with 1994 225 106172 0.64 0.26 fewer parameters, such as are used to a) Blomasses for 1983-90 are average of survey and effectively manage hake stocks elsewhere in catch-at-age model estimates from Methot and Dom (1995); biomass for 1991-94 are derived from an the world (e.g. Payne and Punt 1995), been unciteable PSARC document dated 1995, and is based on compared with the heavily parametrized acoustic data, revised upwards within the catch at age model's range for years where the unsurveyed area stock synthesis approach. problem (see text) is explicitly acknowledged; b) Catch (t), from official statistics, excludes some discards and Consequently, the estimates of Canadian does not include estimates of unreported catch; c) from F == In(l-exp'(catchlbiomass)); in year-I; d) from Dom hake biomass that I have listed in Table 14 (1992); yea{l. are not identical to those appearing elsewhere. I have calculated them by a there is a ''Table 10" showing the variety ofmeans from the published material "utilization percent" (for US and Canada) (see table footnotes) taking into account that derives from the model. This represents many comments made in the most recent the portion of the estimated adult (3+) stock documents. My values have, therefore, biomass that has been caught in each year likely wide confidence limits, at least (equivalent F values are not given). Clearly, ±50%. reported catch divided by this proportion back-calculates the value of the adult stock Some trawl survey biomass estimates that biomass that must have been output by the are independent of the stock assessment model. Unfortunately, these values bear little work have been published for the La relation to biomass figures published in Perouse Bank area, a small part of the area earlier years, or to biomass values cited covered here (48.5° to 49° N, 125.5° to 126° elsewhere in the same document. To W). These range from 179,000 to 439,000 t compound these difficulties, acoustic survey (average 262,000 t; CV = 60% since 1983), results are now said to have consistently depending on temperature (Ware and underestimated hake biomass because hake MacFarlane 1995). Divided by the area of also occur in unsurveyed areas. La Perouse Bank, these correspond to approximately 30 to 74 t of hake km-2 2 Therefore, the confident message conveyed (average 44 t·km- ) and are said to average by Methot and Dorn (1995) has been eroded, 61 % of the pelagic fish biomass. The values and, despite the sophistication of the correlate significantly, but with high modelling, the impression from this material variance, with biomass in this area is of worrying uncertainties entering the calculated from the stock assessment model.

48 Unfortunately, there are no published 250,000 tonnes ofhake, roughly equal to the estimates of the size of this resident stock, human harvest of hake (Livingston and which is not fished in winter. The only Bailey 1985). relevant data are from USSR vessels in 1968, the only year in which year-round It is interesting that the F estimates in Table fishing was recorded (Beamish and 14 for years when Canadian hake biomass is MacFarlane 1985, Table 2). Catch from low are higher than in years when it is high, October through March is recorded as 2.6% suggesting that management does not adjust of the 40,280 t recorded from April through catch to match the climate-driven northward September. Some of this catch would have extension of this stock. But, since in this been from the migratory stock which is still stock, on average, F is approximately equal present during October (0.35% of the winter to M, and in a cooler year more of the stock catch from the 1971 and 1972 USSR catch), remains in US waters, this should not so the figure adjusts downward to 1.7%. On compromise sustainability in this long-lived the other hand, USSR effort was likely not fish. Catches have expanded recently as the as high during the winter months, and so I industry and the managers' response to the suggest using a figure of 2% of the average recent "discovery" of previously unreported migratory stock biomass for the resident hake concentrations. offshore Canadian Pacific hake population, i.e., a biomass ofabout 5000 1. Total mortality, here considered equivalent to a PIB ratio (Allen 1971) can be obtained by adding the F and M values in Table 14; Fishing and natural mortality the most recent figure is Z = 0.91 year--I. Although preceding years suggest lower Fishing mortality was estimated for the figures, several factors argue for a higher Canadian fishery as In( I- catch/biomass) in value. The hake stock in both US and Table 14, and will differ from values that Canada is declining, natural mortality is could be estimated using the entire increasing, the catch data do not include all international catch and the corresponding discards, and no allowance has been made biomass. This perhaps should be done when for unreported catches. Hence my best revising the model presented below. The estimate of current Z for Pacific hake in winter F on the resident offshore hake stock Canadian waters is 0.95 year-I. If in Canada is effectively zero. uncertainties are taken account ofin Ecopath analysis (see Walters, this vol.), then I Natural mortality rate (M) between ages 3 to suggest using the range 0.4 to 1.1 year-I. 12 was estimated as 0.24 year- t by Dorn (1992), who used the relative strengths of "boom" cohorts between successive acoustic Diet composition and/ood consumption surveys. M is considered to be rising as result of increasing abundance of marine There have been 6 published studies on the mammals, which are the principal hake diet of Pacific hake in Canada since that of predators in the northern part of its range Outram and Haegele (1972). The most (see Trites and Heise, this vol.) Over the thorough is that of Tanasichuk et al. (1991), Pacific hake's entire range, the annual while the most helpful for our purposes is consumption by marine mammals is about that based on La Perouse Bank, southwest of

49 Table 15 Summer diet of Pacific hake, by size hake diet offshore. Furthermore, Rextstad (% weight) based on 466 stomachs sampled and Pikitch (1986) demonstrated that diet in August 1983 off the Washington and of Pacific hake may shift dramatically Oregon coast (from Rextstad and Pikitch 1986, Table 2). away from krill to fish with size (Table 15). Hake size"' 40 45 50 55 >55 oeD' Euphausiids"j 93.6 90.9 21.9 20.7 7.9 74.3 Despite the large sample sizes in the La Decapods 3.\ 4.5 13.7 26.5 5.3 5.9 Perouse study, the diet data discussed Herring - 3.8 34.7 28.4 - 7.4 above raises some problems. The field Myctophids I.5 - -- - 0.3 data suggests that the foraging strategy of Eulachon / other smelt - 0.6 23 2.9 - 3.2 hake, an opportunistic ambush predator, Anchovy - - 1.5 2.7 - 0.4 may lead to great variation in the diet Flatfish - - 0.\ \4.5 3.4 1.5 Gadoids - - 1.2 2.4 83.6 5.8 from location to location and from year to Other fish 1.8 0.3 3.6 1.8 0.\ 1.1 year (see also Dill 1983). Depending a) Upper class limit, In cm; upon availability and size, hake may b) Diet over all size classes, assuming a total mortality rate switch from feeding upon euphausiids to of0.36 year-I for 1983 (see Table 14); c) this group contributes 100% ofthe diet in hake < 35 em. decapods to smelt to herring. The area of study of the model below is much larger Vancouver Island (Ware and MacFarlane than La Perouse Bank, and so it may be 1995). There do not appear to be published unwise to use the diet data from that location data on the diet ofPacific hake in winter. without modification. I have therefore adjusted the published figures as follows. In six years of research surveys on La Perouse bank, from August 1983, and 1985 Table 16 Summer diet composition to 1989, 12430 hake stomachs containing (%, by weight) of hake, west coast of food were analyzed (Tanasichuk et al. 1991). Vancouver Island. Diet by weight consisted of 64% Group DC euphausiids (CV: 24%), and 30% herring Euphausids 69.1 (CV: 55%). Decapods 2.9 Herring 18.5 But in the same study in August 1988, Gadoids 2.9 samples from commercial fishing vessels on Eulachon / other smelt 1.6 La Perouse, presumably catching larger hake Myctophids 0.2 and fish on the shelf break, gave only 9% Flatfish 0.8 herring and 82% euphausiids at the same Anchovy 0.2 time as that year's research survey gave 36% Other fish 0.5 herring and 62% euphausiids. Moreover, Outram and Haegeke (1972) recorded 26% If the Rextstad and Pikitch diet composition of 1196 hake stomachs as containing figures are adjusted for the total mortality sandlance (Ammodytes hexapterus), 5% with rate estimated for that year (Z = 0.32 year-\ eulachon (Thaleichthys pacificus), 3% the overall percentage of the diet by item lantern fish (Myctophidae), and 3% pink will be as given in the rightmost column of shrimp (Pandalus spp). Although the Table 15: 74% krill, 5.9% shrimp, 7.4% principal krill species consumed is herring, 5.8% gadoids and 3.2.% smelts Thysannoessa spinifera, Livingston (1983) (including eulachon) for the hake population showed that Euphausia pacifica increases in studied by Rextstad and Pikitch. Although

50 their study was based further to the south toothed whales, and killer whales (resident than our area of interest, many of the same and transient). Estimates of mean body species are known from commercial catches weight for males and females of each to be abundant in Canadian waters. My species were obtained from Trites and Pauly suggestion therefore is to combine these (in prep.). Population estimates were values with those of the La Perouse study obtained from published sources or represent (Table 16). The result is about 70% krill and educated guesses based on the best available about 15% herring with significant amounts information, such as Northridge's (1991) ofother fish and shrimps. global population estimates, scaled down to our reference area. Unless otherwise stated, A second problem arises from the variable mean body weights, food consumption and diet of hake: when we incorporate the diet composition were estimated, for each estimated diet composition into an Ecopath species, as described above for the marine model, small changes in hake diet can have a mammal ofthe Alaska Gyre. high leverage on the balance of the model, because the optional preys have very Diet composition for each ofthe four groups different trophic roles in the ecosystem. equaled the mean diet of the species within When searching for values that balance the the group, weighted by the relative model, a mean diet composition may not population abundance and ration. Summer express the actual shifts in diet that would and winter data for the marine mammal result. species considered here are summarized in Appendix 1, Tables F to I. Using mean stomach fullness data related to time of day, Tanasichuk et al. (1991) Additional information on the assumptions estimate that hake on La Perouse Bank in and estimates used for each species of summer consume about 1.6% of their body marine mammal in the B.C. shelf are as weight per day (i.e., Q/B = 5.84 yea{\ follows. Similar values were calculated using several alternative models for estimating daily ration size, and differential digestion rates Pinnipeds for crustaceans and fish were not thought to Harbour seals, Steller sea lions, northern fur be important. Hake on the shelf edge seals and northern elephant seals occur in consuming more euphausiids had slightly the B.C. shelf in both summer and winter. A lower daily rations. There are no published fifth species, the California sea lion, feeds in winter values. B.C. waters only during winter. The maximum rate of population growth for Marine Mammals northern fur seals and other pinnipeds is generally believed to be about 12% (Small and DeMaster 1995). The PIB ratio was (Andrew Trites and Kathy Heise) therefore set at 6%.

The 13 species of marine mammals that feed Northern fur seals (Callorhinus ursinus) on the British Columbia shelf during from the Pribilof Islands numbered summer and winter were grouped into four 1,019,192 animals in 1994 (Small and categories: pinnipeds, baleen whales, DeMaster 1995). This population breeds in

51 the Bering sea and migrates annually Island. It was assumed that 20% of the male through B.C. waters from January to June on population spent up to one month in B.C. their return trip from California to the waters in the summer, and another month Pribilof Islands (Bigg 1990). Between 3 and feeding here in winter. Given the current 14% of the population feed over the B.C. population estimate of 127,000 elephant shelf (14% January, 6% February, 12.5% seals (Stewart et al. 1994) and an assumed March, 20% April, 12.5% May, and 3.3% sex ratio of 50%, approximately 2,000 June, calculated from the number of fur animals should be present in each of the 6 seals sighted in all areas of the north Pacific months of summer and winter. As there is during pelagic surveys, as given in Figure 8 no dietary information specific to the B.C. ofBigg 1990). This represents an average of shelf, diet composition estimates were taken 5% of the total population over the 6 winter from elephant seals sampled in California months (51,000 fur seals per month) and 6% (Antonellis et al. 1984), i.e., 40% small over the 6 summer months (61,000 per squid, 20% dogfish, 10% rockfish, 10% month). The ratio of males to females was sablefish, 10% hake and 10% miscellaneous assumed to be 0.2:0.8 in both seasons. Mean demersal fish. body weights were for males (30.2 kg) and females (25.3 kg). Steller sea lions (Eumetopias jubatus) breed on offshore rocks and islands along the B.C. Dietary information is based on stomach coast. They generally feed within 20 km of contents from fur seals shot at sea from 1956 the shore during summer, but may venture to 1972 (Perez and Bigg 1981, 1986). The several hundred kilometers away from it British Columbia shelf corresponds to Area during winter. The current B.C. estimate is 9 of the fur seal pelagic survey (Table 11, 9,400 sea lions (P. Olesiuk, pers. comm.). Perez and Bigg 1981) where, in the winter Summer and winter diets were based on scat (Jan-MaL), the animals predominantly eat samples collected at Forrester Island in herring (36%), squid (21 %), rockfish (12%), Southeast Alaska, an area similar to the B.C. salmon (11 %), smelts (7%), sablefish (6%), shelf (Trites and Calkins, unpubl. data). miscellaneous clupeids (5%), hake (1%), Summer diet consists predominately of and sandlance (1%). In summer (April to salmon (17%), herring (16%), pollock June), the animals consume mainly herring (12%), cod (12%), rockfish (11 %) flatfish (52%), salmon (12%), squid (14%), rockfish (11 %), sandlance (8%), skates (3%), octopus (7%), sablefish (4%), hake (3%), sandlance 2%), squid (1%), smelt (1%) and hake (1%) (3%), smelts (1 %), miscellaneous clupeids and other fishes (4%). Winter diet consists (2%) and large pelagic fish such as pollock ofpollock (22%), cod (22%), herring (13%), (2%). salmon (11 %), flatfish (8%), rockfish (5%), skates (5%), sandlance (4%), octopus (4%), Northern elephant seals (Mirounga squid (2%), smelt (2%), sculpins (1 %) and angustirostris) make biannual migrations other fishes (2%). from the breeding beaches in California to deep waters of the Gulf of Alaska (DeLong Harbour seals (Phoca vitulina) are the most et al. 1992; LeBoeuf 1994; Stewart and abundant resident pinnipeds in British DeLong 1994, 1995). Males travel further Columbia. Approximately 40,000 of the north than females and are often sighted 135,000 B.C. population inhabit the outer onshore along the west coast of Vancouver shelf regions (Olesiuk et al. 1990b). Diet is

52 not known in this region and was assumed to to feed for 1 month. Given a total population be similar to the non-estuary waters of of 21,000 whales, a summer population of Georgia Strait (from Olesiuk et al. 1990c). 1167 animals was estimated (range 1000­ Thus, the summer diet comprised 65% 1500). In winter, animals returning from the gadoids (mostly hake), 15% herring, 4% feeding grounds in the Bering Sea have thick midshipmen, 3% salmon, 2% smelts, 2 % blubber layers (Tomilin 1957) and are less sandlance, 2% squid, 1% hexagrammids, likely to feed. It was therefore assumed that and 6% other fishes. The winter diet was only half of the summer population stops to comprised of60% herring, 20% gadoids, 6% feed off the Vancouver Islands shelf in hexagrammids, 4% plainfin midshipmen, winter (585 animals, range 200-1000). 3% salmon, 2% squid, 1% smelts and 4% other fish (based on Figure 22 in Olesiuk et Gray whales were assumed to consume 4% al. 1990c). of their body weight per day based on summer feeding rates for other species California sea lions (Zalophus (Lockyer 1981 b), while the winter californianus) breed along the coasts of consumption rate was based on Innes et al. (Alta) California, US, and Baja California, (1987). Based on Wolman (1985), the Mexico. Each winter (from November to summer and winter diet of gray whales was May) approximately 3,500 males enter B.C. assumed to comprise 90% amphipods, 5% waters. About half feed in Georgia Strait and polychaetes and 5% molluscs. half on the outer coast of Vancouver Island. The diet is not known, but was assumed to Humpback whales (Megaptera be intermediate between that of harbour novaeangliae) migrate seasonally between seals and that ofSteller sea lions. the warm water breeding grounds in winter and rich feeding grounds further north in summer. British Columbia appears to be an Baleen Whales approximate geographic boundary between feeding from Alaska and California (Small Gray, humpback and minke whales occur off and DeMaster 1995). Green et al. (1992) the west coast of Vancouver Island in estimated that 67 humpbacks were present summer and winter. With the exception of year round off the coast of Washington and minke whales, most ofthe baleen whales are Oregon. For modelling purposes, it was migratory and do not feed extensively while assumed that 100 whales were present year transiting the area. However, a few round offthe west coast ofVancouver Island individuals do remain year round in the area. (allowing for a range between 50 and 200 Their production was estimated to be half of animals). While this may appear to be an the 4% maximum rate of population increase overestimate, it accounts for animals that for cetaceans (from Reilly and Barlow migrate through the area and feed while 1986). passing by.

Gray whales (Eschristius robustus) migrate Humpback whales were assumed to off the west coast of Vancouver Island en consume 4% of their body weight per day, route to feeding grounds in the Bering Sea based on summer feeding rates for other (Small and DeMaster 1995). It was assumed species (Lockyer 1981 b). Diet was estimated that one in three animals remains in the area from 30 historical whaling records compiled

53 from animals harvested off the west coast of of Vancouver Island. However, Green et al. Vancouver Island (Nichol and Heise 1992). (1992) estimated 2,100 off the coast of Humpbacks feed primarily on euphausiids California and Washington, with no (80%), herring (10%) and copepods (10%). significant difference in densities between winter and summer. For modeling purposes Minke whale (Balaenoptera acutorostrata) the year round population of Dall's abundance is poorly understood in the porpoises was assumed to be 1,000 (range: nearshore waters of British Columbia. 300-3,000) for the shelf. The year round diet Minimal stomach content data is available as consists of herring (40%), sandlance (30%), the species was not commercially harvested. and small squid (30%) (from Stroud et al. Based on Stewart and Leatherwood (1985), 1981; Jefferson 1990). and the observations of Hoelzel et al. (1989), it was estimated that minke whales Harbour porpoise (Phocoena phocoena) consumed a mix of euphausiids (30%), population estimates along the shelf of copepods (30%), herring (20%) and Vancouver Island were not available, sandlance (20%). There is little evidence to although surveys have been conducted off suggest seasonal movements of minke the coasts of Washington and Oregon whales. (Green et al. 1992). For modelling purposes, the number of harbour porpoises along the The number of minke whales on the shelf shelf of Vancouver Island was estimated at was set at 100, with a range from 50 to 300, 1,000 (range 150-1500), based on sighting and was assumed to be constant year-round. data in Green et al. (1992), and Osborne et This is a best guess in the absence of al. (1988). The diet is similar to that of abundance estimates to extrapolate from. Dall's porpoises and includes herring (40%), Minke whale sightings were rare in aerial sandlance (30%), small squid (20%) and surveys off the Washington and Oregon miscellaneous demersals (10%) (Treacy coast (Green et al. 1992). 1985, Gerrin and Johnson 1990).

Pacific white-sided dolphins Toothed Whales, Dolphins and Porpoises (Lagenorhynchus obliquidens) have not Dall's and harbour porpoises are found in been censused off the coast of Vancouver the shelf area during all months of the year. Island. In Washington and Oregon, these Pacific white-sided dolphins and northern dolphins can be seen in nearshore areas right-whale dolphins are not generally seen (Green et al. 1992). Off the west coast of in nearshore areas, but are found 15-20 km Vancouver Island, they are most frequently offshore and beyond. Resident and offshore seen in July through September at distances killer whales are present year round. The of greater than 10 miles offshore. Total maximum rate of population growth for all population was estimated at 2,000 (range cetaceans is believed to be 4% (Reilly and 1,000-3,000). Based on stomach content Barlow 1986) and so the P/B ratio was data in Stroud et al. (1981) and Walker et al. (1984) diet composition was estimated at estimated to be 2% (half ofrmax)' herring (40%), sandlance (30%), transient Dall's porpoises (Phocoenoides dalli) have salmon (10%), rockfish (10%) and squid not been reliably censused offthe west coast (10%).

54 Northern right whale dolphins (Lisso­ higher than the consumption rate predicted de/phis borealis) are relatively rare on the by the equation of Innes et al. (1987). The shelf, but have been observed occasionally diet estimates for resident and offshore off La Perouse Bank. Recent sightings (Rod whales were pooled from stomach content Palm, Strawberry Research, pers. comm.) analyses in Barrett-Lennard et al. (1995). It suggest that 100 animals (range 50-300) can was assumed that fish-eating killer whales be found offthe west coast year round, often consume 65% resident salmon, 15% shark, in association with Pacific white-sided 15% transient salmon and 5% herring. dolphins. The diet consists of herring (50%), sandlance (40%) and small squid (10%), based on stomach content analyses of Transient Killer Whales animals caught in the squid drift-net fishery The total transient killer whales population (Walker and Jones 1993). in British Columbia is estimated at 170 Resident and offshore killer whales (Barrett-Lennard et al. 1995). As the study (Orcinus orca) can be found along the area comprises about 1/5 ofthe known range Vancouver Island shelf, but they have been of transients, a year round population of 34 studied most thoroughly in the waters of the animals was assumed. Production and sex Inside Passage and in the Queen Charlotte ratio estimates were assumed to be the same Islands, respectively (Bigg et al. 1990, Ford as for resident killer whales. et al. 1994). 290 resident whales and 200 Barrett-Lennard et al. (1995) adjusted killer offshore whales have been identified to date whale consumption rates for transients by (Ford et al. 1994). Given that not all animals incorporating the higher calorific value of are on the west coast of Vancouver Island at one time, a year-round population of 100 marine mammal prey. Transient killer whales consume 73 kg of marine mammals residents and 100 offshores was assumed. 01esiuk et al. (l990a) estimated a production per day, and the difference between male rate of 2.9% for resident killer whales in and female consumption rates was not significant. The diet was assumed to be 75% British Columbia. Small and DeMaster pinnipeds, 20% porpoises and 5% baleen (1995) use r = 4%, a value that is max whales, based on stomach contents collected considered conservative for most cetaceans off British Columbia and Alaska (Barrett­ (Reilly and Barlow 1986. A PIB ratio of 2% Lennard et al. 1995). The diet was assumed of resident killer whales was assumed for to remain constant year round. modelling purposes. Mean body weights were 2587 kg for males and 1973 kg for females (Trites and Pauly in prep.). Barrett­ Lennard et al. (1995) calculated the sex ratio Seabirds of the southern B.C. shelf of female to males to be 0.64:0.36, based on data in Bigg et al. (1990). (Yoshihiko Wada and John Kelson)

Based on the age structure of the resident The complex coastal physiography of killer whale population and the calorific British Columbia creates highly diversified value of prey, Barrett-Lennard et al. (1995) coastal habitats, with a rich marine avifauna, estimated that male and female killer whales but changes in water temperatures during consume 84-85 kg per day, which is slightly ENSO events affect seabirds in the southern

55 portion of this area, due to shifts in prey Palmer (1962). Diet composition were taken abundances and distribution. Here, the from Nilsson and Nilsson (1976). Group estuarine and primarily nearshore species of averages were used for the body weights and seabirds and waterfowl (listed in Palmer's diet composition ofmergansers, and gulls. Handbook of North American Birds (1962 ­ 1988) are not included, as the reference area Table 17 summarizes the population and does not include shallow waters (see above). food consumption estimates, while Table 18 Thus, diet components originating from summarizes the diet composition of seabirds nearshore habitats must be treated here as on the southern B.C. shelf. "imports" to the ecosystem represented by Overall, these tables lead to an estimated the model below. summer biomass of 276 t, a Q/B value of Summer values, (April to September) were 112 year-I, and a winter biomass of 38.5, taken from Vermeer et al. (1983). Sooty with Q/B = 94.8 year-I. Winter population shearwaters, which were present in large numbers were estimated to be 5% of numbers only in May, were included at 1/6 summer populations. Biomass in winter is of their May abundance (Vermeer et al. 14% of that in summer, because it is the 1983). For winter values, average population bigger birds which tend to remain in the values from Pacific Rim National Park of summer areas. Due to the inverse October to March were extrapolated to the relationship between Q/B and body weight, entire area (Hatler et al. 1978). this leads to total food consumption being, in winter, only 12% ofthe summer value. Diet composition and consumption were assumed to be the same in summer and The P/B ratio for "seabirds" as a whole may winter for the species present in both be set at 0.1 year-I, based on data in Muck seasons. Body weight values were obtained and Pauly (1987). mainly from Campbell et al. (1990), and

Table 17 Summer and winter population and food consumption ofseabirds, southern D.C.

W Q Summer pop. Summer Cons. Winter Pop. Winter Cons. (kg) (kg/year) (N) (kg/year) (N) (kg/year) gulls 1.16 109.61 17,786 1,949,398 4,373 479,272 Cassin's auklet 0.15 19.26 738,476 14,225,196 23,822 458,877 Rhinocerous auklet 0.5 53.60 41,621 2,230,894 245 13,123 Tufted puffin 0.6 62.59 56,743 3,551,289 -- Common murre 0.8 79.92 8,139 650,522 897 71,690 Marbled murrelet 0.4 44.34 8,000 354,720 203 8,980 Pigeon guillemot 0.4 44.34 2,721 120,670 473 20,961 Merganser 1.2 112.81 4,000 451,241 3,525 397,704 Pelagic cormorants 2 174.15 2,801 487,824 71 12,350 Leach's storm petrel 0.05 7.57 38,333 290,230 551 4,172 F.-tailed storm petrel 0.05 7.57 1,765 13,365 - - Northern fulmar 0.78 78.22 200 15,644 92 7,220 Sooty shearwater 0.68 69.61 83,333 5,800,907 5,263 366,373 Db.-crest cormorant 1.9 166.72 4,000 666,876 10,167 1,694,977 Common loon 3.3 266.55 600 159,931 314 83,709 grebes 1.24 116.00 300 34,800 300 34,800 Sum 1,008,818 31,003,506 50,295 3,654,208

56 Table 18 Summer and winter diet composition ofseabirds, southern D.C.

Predator\ Prey Benthic Small Carn S. herb. L. herb Small Velella Small a org. ) pelagics zoopl. zoopl. zoopI. squids velella salmon Gulls 50 30 10 10 Cassin's auklet 10 35 35 15 5 Rhinocerous auklet 83 15 2 Tufted puffin 83 15 2 Common murrelet 60 40 Marbled murrelet 80 15 5 Pigeon guillemot 30 70 Merganser 35 65 Pelagic cormorants 26 74 Leach's storm petrel 6 25 55 5 4 5 Fork-tailed storm petrel 6 25 55 5 4 5 Northern fulmar 100 Sooty shearwater 73 27 Double-crest cormorant 10 90 Common loon 30 70 Grebes 10 90 Summer diet (%)0/ 4.1 42.4 17.2 16.1 7.9 10.2 -0.1 2.0 Winter diet (%r' 12.2 62.9 5.8 4.4 2.7 3.6 -0.0 8.4 a) Consisting ofcrabs and other invertebrate organisms, and imports from the intertidal; b) Weighted average for all birds combined. Based on diet composition from this table, and Q/B and other statistics from Table) 7.

Fisheries catches Data Sources and Methods

The data covers the nine year period from (Eny A. Buchary) 1985 to 1993 and were extracted from the B. C. Commercial Catch Statistics Database Commercial landings are useful when of the Department of Fisheries and Oceans constructing ecosystem models, as they tend (DFO, Vancouver Office). Data categories to be estimated more accurately than are include year, type of gear, species and other fluxes. However, some knowledge of weight (kg), by species. These categories are the biology of the fishes generating the important when corrections have to be made landings is important for these to be from dressed (with the head either on, or off) correctly interpreted, as is a knowledge of to round weight. A QuickBasic programs discarding practices. Exclusion of some was used to assemble and extract data from salmonid species to avoid overestimation of the diskette kindly provided by DFO. export from the system is a good example. The fact that DFO statistical areas are open For DFO's purposes, British Columbia offshore, while the model area is defined waters are divided into northern and vertically by the 300 m bathymetric contour southern halves, further divided into (see above) is another related problem, here Statistical Areas whose boundaries are often assumed to be to be negligible, as the bulk defined by a headland or island (Capt. G. ofthe landings stem from shallower waters. Nelson, pers. comm.), and which have open

57 seaward boundaries. The reference area for a breakdown by species, i.e., identifies the model of southern B.C., as defined above, species included in each functional group. thus consist of DFO Statistical Areas 11 (111), and 21 (121) to 27 (127). As might be seen, landings vary very strongly between summer and winter, with All landings are presented for two six-month catches of salmon, sablefish, hake and periods ("Summer", from April to halibut being higher in summer, while September and "Winter", from October to urchins and miscellaneous catches, March), and are reported as "round weight", especially of roe herring, are higher in i.e., the weight ofthe whole fish as it comes winter. The landings in Table 19 from the water before any treatment or underestimate true catches, as they do dressing (Anon 1992). A conversion factor almost anywhere (Alverson et al. 1994). of 1.18 was used to convert salmonids Table 20 suggests that in British Columbia, caught by troll and troll freezer from dressed discards represent on the average, in target weight head-on to round weight. A conver­ species, 22.3% oflandings. sion factor of 1.33 was used to convert halibut from dressed head-off to round Regarding non-target species, Alverson et al weight. All other catches are reported as (1994, Table 10) indicate a ratio of round weight. Salmonid landings from the discarded by-catch to landed catch of 2.21 southern B.C. Shelf include red and white for "British Columbia Cod Trawl", and this chinook, coho, chum, pink and sockeye, but figure may be used for the demersal group in only chinook and coho are permanent the model. residents. Chum, pink and sockeye, on the l 2 other hand swim through the area, without Table 19 Mean landings (kg year- .km- ) feeding, on their spawning migration. from the southern shelf of B.C., by Hence, they are not considered to belong to functional group and season (based on DFO the ecosystem. DFO data also record salmon data for the years 1985 to 1993). roe, which was not considered here as it stems mainly from chum salmon. Functional group Summer Winter Bivalves 71.8 72.7 Hake are problematic because the DFO data Miscellaneous benthos 1.3 0.6 do not include the apparently high volumes Sea cucumber/urchins 8.7 33.8 that are traded at sea (T. Pitcher, pers. Octopus 0.6 0.5 Squids 4.8 0.2 comrn.). Discarded by-catches are discussed Planktons 0 0.03 below. Cods/black cods 346.6 142.1 Prawns/shrimps/crabs 118.7 33.4 Large pelagics 1303.4 143.8 Results and Discussion Small pelagics 1.9 0.3 Miscellaneous demersal 177.4 87.1 Overall, an average of about 75 000 t of fish Ocean perches/rockfishes 377.3 401.1 and invertebrates are reported to be landed Sharks 71.5 68.4 annually from our 30,000 km2 reference Transient salmon 419.4 252.7 2 Resident salmon 454.7 4.3 area, or 2.5 t·km- • A breakdown by functional groups and season is presented in Other 8.0 408.6 All landings 3366.2 1649.6 Table 14, while Appendix 1, Table I gives a

58 Acknowledgments Balancing the Southern B.C. Shelf Model I am grateful to Leonardo Huato for the QuickBasic programs used for extraction of the data. I also thank Dr. Carl Walters for his (Judson Venier) comments on salmonid biology and Dr. Tony Pitcher for his input on hake catches. The following is a brief account of the Lastly, I would like to extend my special actions taken to balance th.e Ecopath model thanks to Ms. Maureen Kostner (Catch and of the self of southern B.C., by functional Effort Unit, DFO Pacific Region), Ms. Maria group (See also Table 20 and Appendix 1, Poon (Statistics Unit, DFO Pacific Region) Table J).). and Captain Gordon Nelson (Head of Operation and Radio Room, DFO Pacific Marine Mammals Region), for all their help in providing data and information. The data from these groups were considered to be the most reliable and thus all estimates remained unchanged.

Table 20 Details on trawl catches in B.C. waters, allowing estimation of observer bias (ratio of at sea I offload weight), and of discarding (ratio discarded I kept weight). Based on hauls performed by 84 trawlers from February 16 to May 17, 1996 (Source: Archipelago Marine Research, Victoria, B.C.).

Retained Ratio Discarded at Sea Ratio Functional At sea Offload at sea I Marketable Not discarded! groups estimate weight offload Dead Live marketable kept weie:bt weigbt Yellowtail 3,641,512 3,758,515 0.97 1,192 0 18,139 0.005 Widow 991,371 1,214,935 0.82 6,126 0 215 0.006 Canary 253,931 216,295 1.17 419 0 415 0.003 Silvergrev 619,926 659,023 0.94 567 0 2,846 0.006 Yellowmouth 2,342,587 2,169,677 1.08 809 0 51,725 0.022 Rougheye 721,359 654,578 1.10 1,560 0 2,462 0.006 Shortraker 104,441 71,574 1.46 4 0 158 0.002 Redstripe 973,000 986,627 0.99 1,458 0 176,614 0.183 Sharpchin 193,745 284,903 0.68 667 0 122,452 0.635 Sablefish 83,849 82,657 1.01 9,767 38,281 99,184 1.756 Pacific cod 294,190 284,489 1.03 282 507 15,123 0.054 Dover 1,730,936 1,814,032 0.95 1,490 9,535 103,155 0.066 Rock 528,371 499,200 1.06 25 225 87,821 0.167 Lemon 228,110 185,672 1.23 794 2,331 89,415 0.406 Petrale Sole 278,317 266,094 1.05 70 445 9,363 0.035 Lingcod 394,876 556,715 0.71 408 3,669 11,795 0.040 Pollock 993,280 1,084,001 0.92 0 0 91,140 0.092 Hake 7,429 4,618 1.61 4 0 77,758 10.467 Dogfish 140,379 136,645 1.03 0 0 1,205,445 8.587 Turbot 2,030,668 1,709,479 1.19 0 100 1,440,253 0.709 Skate 148,610 257,584 0.58 0 0 246,501 1.659 All Species * 21,385,504 21,125,047 1.01 32,175 55,795 4,676,474 0.223 *= Including species other than those hsted 10 thiS table 59 Birds on the Pacific cod group to reduce pressure on it, and to switch a small percentage to Only one change was made to the input their own group. The same was true for the parameters for this group as all other herring/small pelagic group, whose biomass estimates were deemed reliable. The DC was could not support the pressure exerted on it. reduced from 0.041 on decapods to 0.040 and the remaining 0.1% was put into sea stars. This was done to reduce predation Salps pressure on the decapod group and also to account for some sea birds predation on sea Inputs remained unchanged. stars (pers. obs.). Copepods Spiny Dogfish Same as for chaetognaths (see below). This box includes salmon (both resident and transient) and miscellaneous pelagic fish Phytoplankton such as walleye pollock. P/B was increased t from 0.1 (Polovina, this vol.) to 0.75 year- Biomass was estimated by dividing the to account for other pelagics incorporated primary production of the southern British into the box. Biomass was decreased, for the Columbia shelf area (Pauly, this vol.) by the same reason, from 1.75 (Polovina, this vol.) P/B of primary producers from the Strait of 2 to 5 t·km- • Georgia model (Mackinson, this voL). Detritus biomass was estimated using the empirical equation in Pauly et al (l993b; see Pacific Halibut also Venier, this voL). All input parameters remained the same. Pacific cod/ miscellaneousfishes Pacific Hake I reduced predation of this group from This group proved to be a major problem sablefish to hake. The DC of cod was also when balancing the model, for its biomass is changed, from a highly specialized diet of very high. The amount of predation pressure small pelagic fish, to a mix of benthic it exerts on anything it eats, no matter how organisms, both to reduce pressure on the trivial a percentage ofthe diet, was large and herring/pelagic fish group and to reflect the posed the greatest difficulties. Since many diversity of feeding strategies of the various hake inhabit the system, and the Southern demersal fishes included along with the cod. British Columbia Shelf is so open, it was necessary to reduce DC of hake on all Juvenile andAdultSableflSh components ofits diet by half and to assume that 50% of its diet is an import. Still, it Estimates for these two groups remained the was necessary to lower the diet composition same.

60 Table 21 Basic parameters ofsouthern B.C. shelf Carnivorous Jellies 2 model. (Biomass in t·km- ; PIB and QIB in year-I.) Input parameters remained unchanged. Group B P/B Q/B EE Transient orcas 0.002 0.20 12.13 0.00 Odoncetae 0.020 0.40 15.59 0.61 AllMacrobenthos Groups Pinnipeds 0.180 0.40 15.33 0.25 Mysticetae 0.750 0.02 13.02 0.08 The data for all of these groups were Seabirds 0.009 0.10 112.00 0.00 scarce, especially regarding diet Spiny dogfish 5.000 0.75 5.00 0.85 compositions, but were used when Pacific halibut 0.473 0.44 1.73 0.54 available. Missing parameter estimates Pacific hake 44.000 0.75 5.84 0.24 were derived that were deemed compa­ Pacific cod 7.000 1.20 4.00 0.78 Juv. sablefish 1.500 0.60 6.60 0.39 tible with the model. Note that here, the Adult sablefish 0.100 0.08 3.73 0.00 "shrimps" group includes cephalopods. Herring/sm. pel. 17.269 2.20 11.00 0.95 Cam. jellies 6.190 7.00 23.33 0.20 Euphausiids Decapods 10.000 1.80 7.20 0.54 Shrimps 5.000 1.20 8.00 0.75 Polychaetes 20.000 3.00 33.33 0.48 Krill are heavily fed upon within this Sea stars 5.000 0.40 4.44 0.56 ecosystem and as a result, their biomass Sea urchins 5.000 0.40 4.44 0.70 was estimated to be very high. The Brittle stars 5.000 1.80 9.00 0.80 original biomass estimate from Jarre­ Bivalves 5.000 0.70 7.78 0.88 Teichmann (this vol.) had to be Amphipods 5.000 2.40 12.00 0.96 disregarded as it was not high enough Euphausiids 29.586 3.70 15.00 0.95 to support the predation pressure upon Chaetognaths 3.000 12.00 40.00 0.46 this group. An EE value of 0.95 was Salps 5.000 3.00 12.00 0.58 Copepods 16.700 55.00 183.33 0.93 entered and the biomass was left Phytoplankton 26.000 135.00 0.00 0.80 unknown. Detritus 7.000 - - 0.64 Chaetognaths Herring / smallpelagicfish Since Q/B was unknown for this group, a In addition to Pacific herring, Pacific GE of 0.3 was entered; the other parameters sandlance, anchovies and sardines are were not changed included in this group. To obtain a reasonable biomass estimate, an EE of 0.95 Results and a GE of 0.2 were entered (Christensen, pers. comm.) and Q/B was left unknown. In Figure 3 presents the flowchart of the model addition, the DC values for this group were based on the basic inputs in Table 21, and altered to reflect a more planktonic diet the diet matrix in Appendix Table J. instead of one which concentrated on euphausiids. This is believed to be correct as I abstain from presenting a detailed analysis herring and the other pelagic fishes in the ofthis model, given its very tentative nature, box are too small to feed heavily on such itself due to the openness of the ecosystem large prey (V. Christensen, pers. comm.). the model is meant to describe.

61 e. '--'Flow

-- Connector --- H-.t -+ Flow 10 deblUI --. RapIrIllon ..:r- , Import

I 'r PInnIpeds

i 4 ~ ..1 .2 .c D­ 0\ o N ~ to- 3

AmphIpods

2 36.0

P~

Figure 3 Flow chart of trophic intesactions in the shelf system off southern British Columbia. All flows are expressed in t.km-2.year- and all biomasses in Hun- ; minor flows are omitted, as are all backflows to the detritus. resulting in strong tidal current. Tides are STRAIT OF GEORGIA mixed, semi-diurnal with a maximum daily MODEL range of 0.5 m (Harrison et a1. 1983). Extinction coefficient (k) and Secchi disc depth (m) range from 0.16 to 0.95, (mean = System definition and primary 0.37) and from 2.69 to 9.78 (mean = 5.9 m), respectively (Stockner et a1. 1979). production The Fraser River flows into the southern end (Steve Mackinson) and creates a large brackish-water surface plume (Legare 1957) which extends far out from the river mouth. The system can thus System definition and oceanographic be viewed as a 'narrow basin intensely mixed conditions at both ends, with a major point source of freshwater in the southern half' (Harrison et The Strait of Georgia is a partly enclosed a1. 1983). basin on the west coast of Canada lying between the mountainous coast range of The Strait is an economically important area mainland British Columbia and the southern (Harrison et a1. 1983) and has been the half of Vancouver Island. Essentially, the object ofmany ecological and oceanographic Strait is a narrow basin intensely mixed at investigations (Arai and Mason 1982). As a both ends, with a major point source of result, there is much literature pertaining freshwater in the southern half The Strait of specifically to the Strait of Georgia. Georgia is a valuable fishing and nursery area However, many of these studies have been for the young of several major stocks of done on single species in isolated inlets, bays, Pacific salmon and for large stocks of and/or fjords and the estimates oflife history herring. It also provides a lesser fishery for parameters which have resulted only apply to cod, groundfish, shellfish, shrimp and crab. a small area of the entire Strait. This makes integration ofdata for various species groups The Strait of Georgia is 200 km long with a difficult. In addition, many studies covered mean width of 30 km, a total area of 6,900 only a limited part ofthe year. 2 km- , and a mean depth of 156 m; it is open to the Pacific Ocean through the Juan de One characteristics of well-defined Fuca Strait to the south and through ecosystems are strong interactions within Johnstone Strait to the north (Harrison et a1. themselves and modest, controlled 1983). Mean summer temperature is 10.90 C interactions with the outside. The Strait of (annual range 50 C). Prevailing winds are SW Georgia ecosystem is very complex, but it in winter and NW in summer. Precipitation fulfills the criterion of being reasonably well ranges from 90 - 200 cm·year"l. Water defined as an ecosystem. Some fish (herring, circulation in the southern half of the Strait salmon) use the Strait only for migration to of Georgia is strongly affected by the large spawning grounds and feed minimally while discharge of the Fraser River, which in the Strait. Thus, although they might contributes to 80% of total land runoff and comprise a large percentage ofthe total fish causes most salinity variation in surface biomass, they impact the system very little. waters. The passages to the Pacific Ocean are generally narrow and restrict tidal flows,

63 biomass, they impact the system very little. (dominant in the Strait), and based on This is discussed in more detail further Olivieri et al (1993, Table 1) may be below. suggested. Allochthonous input to the Georgia Strait is estimated to be as great as total primary production, i.e., about 2.106t Phytoplankton Production year'I (Seki et al. 1969), possibly leading to an eutrophic state (Parsons 1980). The phytoplankton community is dominated by diatoms. Advection, turbulence, zooplankton grazing and summer nitrate Macroalgae (kelp) production depletion collectively impart a hetero­ geneous distribution to phytoplankton in No data on kelp production in the Strait of surface waters. Three estimates of annual Georgia were found in the literature, and Primary Production (PP) were found: hence data from similar areas were used (Table 22). • Parsons et al. (1970): 120 g Cm·2 year·1 ; 2 Table 22 Standing crop (biomass; t km· ) • Stockner et al. (1979): 345 gCm·2 estimates ofLaminaria spp. year;-I and Location Biomass References • Harrison et al. (1970): 280 g Cm·2 Helgoland, Germany 12,700 LUning (1969) year.·1 Nova Scotia, Canada 16,000 Mann (1972) Lagoon Pt., Alaska 14,500 Calvin & Ellis (1978) UJ Assuming 0.1 gC = 0.2 g dry weight N.W. Atlantic 3,685 Brady-Campbell et al. (1984) ~ 1g wet weight (Mathews and Mean ofall locations 11,720 this study Heimdal 1980), the estimates convert a) Converted from dry weight, assuming dry weight = 21 % of wet weight; b) Brady-Campbell et aI. (1984) also estimated an to: average PIB ratio of4.43 year"l, used here.

2 • 1200 t wet weight km· year'I; Assuming that kelp grow on rocky shores to 2 • 3450 t wet weight km- year'I; and depths of about 20 m would suggest, given the bathymetry of the Strait of Georgia • 2800 t wet weight km·2 year"l. (Guenette, this vol.), that kelp cover about 2 % of its surface area, which may be used as The mean of these values is 2483 t wet a raising factor for the contribution of kelp 2 weight km· year' I, and this may be used to total primary production (Table 23). as input.

Several, highly variable estimates of Table 23 Summary statistics on primary Production:Biomass (P/B) ratio were production (wet weight) in the Strait of found in Stockner et al. (1979), Georgia. expressed in terms of Producer PP (t·km'··year" ) P/B (year"') Area (%) photosynthesis/chlorophyll a. For our Phytoplankton 2483 200 98 purpose, a P/B ratio expressed on a wet Macroalgae 52700 4.43 2 weight basis is required, and a value of 200 year'I, pertaining to diatoms

64 Zooplankton (incl. Jellies) Macrobenthos

(Judson Venier and Claudia Oeteau) (Sylvie Guenette)

Table 24 summarizes key information on In the mid 1960s, Ellis (1967a, 1967b, zooplankton and jellies in the Strait of 1968a, 1968b, 1968c) conducted an Georgia. extensive study ofthe benthic community of the Georgia Strait, based on stations Our suggested diet of jellies in summer is: scattered all over the Strait, though at depths herbivorous zooplankton (71 %); carni­ greater than 10m. Stations were sampled (3 vorous/omnivorous zooplankton (26%); and replicates) once or twice during a three year other jellies (3%) (Arai and Jacobs 1980; period (1965 to 1967). Since depth is a Larson 1987a, 1987b; Purcell 1991a, major factor in benthos distribution, these 1991 b), though fish larvae are also taken samples were here grouped into depth strata (Arai and Fulton 1982; Purcell 1989). following Levings et al. (1983, Table 1). (See Table 25). The euphausiacean Euphausia pacifica is one of the most important species of Table 25 Definition of depth strata (m) zooplankton in the Strait of Georgia, for the Strait of Georgia, with estimated representing 50% of the summer biomass areas after Levings et al. (1983). (Heath 1977). Harrison et al. (1983) Depth strata Area (km") Area (%) Neocalanus plumchrus consider to be the 20-50 800 10.8 most abundant component of the 50-100 1,560 21.1 zooplankton, but it occurs in surface waters 100-200 2,130 28.8 (20 - 50 m) only in early spring; in early 200-300 150 21.2 summer, N. plumchrus migrates downward, 300-400 1,000 13.5 with maturation and spawning occurring in >400 330 4.5 deep water (400 m, in the center of the Total 7,390 100 Strait). Samples from intertidal stations in the Puget Sound Table 24 Biomass (wet weight); P/B; Q/B; and P/Q for zooplankton, Strait ofGeorgia. (Nyblade 1979) were used to account for the intertidal zone Biomass P/B Q/B which is important to shore Group (t.km,2) (year,l) (year,l) birds and many invertebrates. 1 Herbivorous zooplankton 7.8'1 8.4° 10.5°} Here, samples were stratified 2_9 CJ lOS) Cam.lomniv.zoopl. (excl. jellies) ---- by sediment type (rock, gravel, Misc.jellyfish and ctenophores 312"J 15 ISO·} sand, and mud) following ,) Probable summer value; yearlt: range is 0.05 - 1.4 g·m,3; Levings et al. (1983, Table 2); b) From Harrison et al. (1983); C From Heath (1977); estimate includes Euphausia pacifica; d) Annual maximum for organisms >350llm at Table 26 shows our depth to 400m: range is 0.1 - 2.0 g.m,3; e) Based on maximum daily intake of herring larvae in Kulleet Bay (Purcell 1989), converted to stratification by bottom type. ash-free dry weight using data on Aequorea victoria in Larson (1986).

65 Table 26 Bottom types in the intertidal weight per day (Mottet 1976). Since only areas of the Strait of Georgia (from 5.5% ofthe population are juveniles (Adkins Levings et al. 1983). et al. 1981), the mean weight of a legal size red sea urchin (S. franciscanus), that is Substrate type Slope t) Area (km') 565g, was used for estimating consumption. Rocky shores, steep 60 54 Rocky shores, gentle 45 266 By this reckoning, an adult would eat 23 Gravel 25 93 times its body weight per year. Comparing Sand 10 424 these values with the relatively low values ~n Mud 5 408 Christensen (1995, Table 4), the more Total --- 1,245 conservative value of 23 was chosen. The Q/B value for carnivorous benthos comes For each class of organism, biomass values from Olivieri et al. (1993). The values of z (in g.m- ) were averaged by stratum (2-7 gross efficiency (P/Q) come from stations per stratum). The intertidal zone, Christensen (1995), while a conservative and those of Ellis' stations located at depths value of ecotrophic efficiency (EE) was <20 m, were each allocated 50% of the chosen following Christensen and Pauly intertidal zone. The final biomass of (1993). Mortality rate estimates (Z; year-I) specimens were averaged by weighing each for commercial species are: red sea urchins stratum by its estimated area. The 0.2 (Jamieson 1984); abalone 0.23-0.91 methodology underestimates large (Breen 1980); geoducks, 0.3-0.4 (Noakes burrowing clams such as geoduck Panope 1992); and Dungeness crabs, 2.5 (Smith and abrupta, abalone Haliotis kamtshatkana and Jamieson 1991). Table 27 summarizes our the crabs e.g. Dungeness crabs Cancer suggested biomass and rate estimates for the magister. Detailed results are presented in two benthos boxes: Appendix 1, Table J. These data were used to identify two Table 27 Tentative estimates of biomass, PIB, QIB, P/Q functional groups: 1) small and EE for the macrobenthos ofthe Strait ofGeorgia. herbivorous or detritivorous Benthos Biomass PIB Q/B Gross eff. EE benthos and 2) large carni­ l J 2 (vear· ) (vear- ) (P/Q) (-) vorous/detritivorous benthos. e:roup (t'km- ) Herb. & detritiv. 375 0.6 23 0.15 0.9 The latter group includes Carnivorous --- 2.5 10 0.2 0.9 starfish and crabs.

Abalone typically eat 10-15% of their Diet composition of adult crabs is 20% weight of seaweed per day (Mottet 1978, young crabs and 80% other benthic based on data from the Queen Charlotte organisms (Stevens et al. 1982, Table 5). Islands). Given a mean weight of 200g for Fifty % of the animals eaten (cannibalism legal size specimens (Breen 1980), and low excepted) consist of carcasses and other recruitment, one specimen would eat 36-55 detritus. This is similar to values used for times its own weight per year. Along the similar benthic communities by Jarre­ California coast, at 15°C, red sea urchins Teichmann et al. (in press). Table 28 gives (Strongylocentrotus purpuratus) eat algae at our suggested diet composition for the a rate corresponding to 6.4% of their body macrobenthos.

66 Table 28 Approximate diet composition of herbivorous/ sources also used for detritivorous benthos, and of adult crabs in the Strait of Table 29, as follows: Georgia. euphausiids (40% by Consumer Macrophytes Detritus Herbivorous Carnivorous weight), copepods (15 group and algae benthos. benthos %); decapods (10%); Herb. & detrivores 40 50 - 10 shiner, sandlance and Adult crabs - 40 40 20 flatfish (10%); and cannibalism (25%), this The Demersal Fish "Box" high value being due mainly to over­ aggregation within the box. These estimates are expected to be much modified when the model is balanced, due to the leverage effect (Judson Venier and John Kelson) discussed in Pitcher (this vol.). No single reference was found which reviewed the status of demersal fishes in the The Pelagic "Box" Strait of Georgia, and Table 29, which presents key parameters for the group, was assembled from a number of disparate sources. (Rik Buckworth)

Overall, the estimated biomass of the The data summaries below refer to the demersal fish box, consisting mainly of pelagic component of the Strait of Georgia -2 hake, is 198,306 t, or 28.7 t'm , while the ecosystem, excluding planktonic elements mean (weighted by biomass) P/B and Q/B and marine mammals, and concentrating ratios for this box are estimated as 0.6 and upon species for which information was 5.54 year-I, respectively. available. Surprisingly, although this component contains species which are The diet composition of the demersal fish subject to intensive and very valuable box was estimated, based mainly on the fisheries, the particular data types upon

Table 29 Biomass, PIB, and Q/B ofdemersal fish in the Strait ofGeorgia.

2 Species a) Biomass (t'km- ) PIB (year-I) Q/B (year-I)b) Hake 1,2) 23.5 0040-0.72 5.84 Walleye Pollock J.') 3.3 0.7-0.9 4.76 C Rockfish ',b) OAO-I.07 ) 0.06-0.28 3044 Lingcod I,~) 0.30-0,51 0040-0.76 3.30 d Flatfish \/, IU) 0.32-0A2 ) 0040-1.15 3.21 Pacific Cod II.IL) 0042 1.03-1.50 3043 Sum (means) 28.74 (0.6) (5.54) a) References for biomasses and PIB ratios: 1) Saunders and McFarlane (1994); 2) Dom et al. (1993); 3) Shaw et al. (1989); 4) Saunders et al. (1989); 5) Yamanaka and Richards (1992); 6) Hand and Richards (1991); 7) Richards and Yamanaka (1992); 8) Smith et al. (1990); 9) Fargo (1995); 10) Fargo (1991); 11) Stocker et al. (1995); 12) Westrheim and Foucher (1987); b) From Table II and empirical relationships in Pauly (1989), except for hake, taken from Pitcher (this vol.); c) Biomass estimates pertain to west coast ofVancouver Island, divided by approximate shelfarea (see Pauly this voL); d) Biomass estimates pertaining to rock and English soles in Hecate Strait, divided by rough estimate ofarea (-30,000 km-\

67 which Ecopath modeling depends were not 100,000 t for the early 1980s down to about easy to come by in the literature. There are 30,000 t (Hay and Fulton 1983; Stocker thus several shortcomings in the data 1993; Environment Canada 1994). Annual presented. However, one of the objectives of spawning stock biomass in the late 1980s for Ecopath is to identify the relative the Strait of Georgia was between about importance of different pieces of 30,000 and 60,000t, depending on year and information and it is hoped these estimation method (Schweigert and Fort shortcomings are not major. 1994). Northcote and Larkin (1989) estimated the numbers of salmonids Some aspects of the pelagic box should be produced by the Fraser River system at clearly emphasized. One is that some about 14 million fish annually. The important components - herring and salmon historical abundance of pink salmon (0. - are not permanent occupants of Strait of gorbuscha) alone was around 48 million fish Georgia. Herring enter the Strait to spawn (Ricker 1989), but has been as low as 2 and are subject to intense fishing and million in 1961. predation. Herring spawn is itself the subject of a fishery and is heavily utilized by other The dominant pelagic species in the Strait of species and may be a very significant input Georgia ecosystem are herring Clupea into the Strait as a stimulus for secondary pallasi; spiny dogfish (Squalus acanthias); production (Hay and Fulton 1983). Because sockeye salmon Oncorhynchus nerka; coho little of the life cycle is spent in the Strait of Oncorhynchus kisutch; chum Oncorhynchus Georgia, probably a fair proportion of the keta; pink Oncorhynchus gorbuscha; herring and salmon biomass should be chinook Oncorhynchus tshawytscha; considered as an import to the system. steelhead Oncorhynchus mykiss; lampreys Lampetra ayresi,' and eulachon Thaleicthys Herring and salmon abundance fluctuate pacificus. Other pelagics of the Strait of considerably interannually. Current Georgia are, in deeper waters, the abundance also reflects overfishing. Herring bathypelagids and the myctophids, not biomass estimates vary between about considered here (but see Keiser 1922).

Table 30 Ecological parameters of key pelagic component ofthe Strait of Georgia.

Groups') Residence Biomass Ration P/B 2 time (t'km- ) (%W dafl) (year-I) Herring I,L,j,O) Nov-April, 20 weeks 6.67 0.60 Dogfish4,C} Resident 1,45 3 0.20 Salmon)' 0, '.a,e) Aug.-Oct., 10 weeks 5.80 5 0.75 Lamprel'" g) Apr.-Sep. ~ 90days. 1.04 - 4.60 a) References: 1) Environment Canada (1994); 2) Hay and Fulton (1983); 3) Stocker (1993); 4) Thompson (1994); 5) Northcote and Larkin (1989); 6) Tutubalin and Chuchukalo (1992); 7) Walters et al. (1978); 8) Beamish and Youson (1987); b) Biomass based on mean of 1985-1989 estimates in Stocker (1993); c) Low risk yield (2,000 t), multiplied by 5 to get biomass: F is low, mean 1985-1990 was 948 t (Thomson 1994); d) Biomass based on annual yield of30,000 t, raised by exploitation rate of75%; e) Ration estimates were 3-10% ofbody weight daily (Tutubalin and Chuchukalo 1992); t) Total biomass declines from 13,000 t to 1,300 during residence (5); g) Life cycle is about 1 year; lampreys each attack 0.8 fish/day, with average size 54g (Beamish and Youson (1987).

68 Table 31 Approximate diet composition of key components ofthe pelagic box ofthe Strait ofGeorgia ecosystem.

Predators·' Preys ( in % volume) Herring .~. 0. zooplankton (90) benthos (10) -- Dogfisho. u •• °1 zooplankton (56) hake (14) herring, shrimp, salmon (15) Salmon"} macro(zoo)plankton, squids, herring, myctophids, other fishes Lampreys '''} Herring (84) Salmon (16) -- a) References: I) Arrhenius and Hansson (1993); 2) Environment Canada (1994); 3) Hay and Fulton (1983); 4) Stocker (1993); 5) Beamish et aL (1992); 6) Jones and Green (1977); 7) Tanasichuk et al. (1991); 8) Thomson (1994); 9)Tutubalin and Chuchukalo (1992).

Table 30 summarizes the biological data for Marine Mammals and Birds the different groups. I have included the group "other" to emphasize that the information presented cannot include all (Yoshihiko Wada) species in the pelagic component, but only those for which significant information is Introduction available. Table 31 lists the prey of the different groups, and Table 32 provides The purpose of this exercise being to numerical catches ofsalmon by species. construct a complete model of the flow of biomass between different trophic levels The herring biomass estimate was based on within the Strait of Georgia in the summer data presented by Stocker (1993; mean of months (June 23 - September 22) of the total biomass estimates in Figure 10, p. 280). 1980s, this contribution provides the Numerical abundance and exploitation information on food consumption and food estimates for salmon species were from composition required for including the Northcote and Larkin (1986). Landings were marine mammals and birds which occur in from Anon. (1987, 1988). Walters et al. the Strait of Georgia during the summer (1978) provided a detailed model of passage months. of juvenile salmonids through the Strait of Georgia. Marine mammals Catches for salmon for the Strait of Georgia 1 are around 30,000 t'year- , but vary a lot The marine mammals found in the Strait of among years, e.g. 1986: 25,000; 1987: Georgia in the summer are killer whales, 39,000 t.

Table 32. Estimates of Salmon Abundance, and catches, in numbers (Northcote and Larkin 1989).

Species Abundance Catch Remarks Pink 5,000,000·) 3,000,000 Georgia, Juan de Fuca & Johnstone Straits Chum 776,000 140,000 Fraser & Juan de Fuca Coho 441,000 303,000 Mean of 1973-82 Sockeye 7,400,000 5,400,000 --- Chinook 104,000 646,000 --- a) May reach 9,000,000+ In odd years.

69 Dall's porpoises, harbour porpoises, and thinks that Dall's porpoises are more harbour seals (see Osborne et al. 1988). plentiful in the Strait ofGeorgia. There are two different groups of killer whales: 1) "residents", which generally stay Thus, midranges of 1,125 and 500 were within the strait, and which eat fish, and 2) retained as best estimates for Dall's and "transients", which travel along the west harbour porpoises, respectively, coast of British Columbia, and which eat marine mammals, such as seals and dolphins The harbour seal population of 14,326 was (Ford et al. 1994). determined from the data in Olesiuk et al (1990, Figure 21), based on field data were Estimates of killer whale population size collected in 1988, a year assumed to be were obtained from a series of interviews representative ofthe 1980s. with marine mammal specialists in B.C, i.e., Andrew Trites, Graeme Ellis, Kathy Heise, Mean body weights were obtained from and Lance Barrett-Lennard. The Trites and Pauly (in prep.) for all species compromise numbers of 80 for resident and except transient killer whales, whose 11 for transient killer whales were derived average body weight of 3,550 kg was taken by averaging the figures provided by these from Barrett-Lennard et al. (1995). specialists. The daily ration ofharbour seals in the Strait Similarly, in the absence of any systematic was from Olesiuk et al. (1990); for all other or comprehensive study, the abundance of species, ration and Q/B were estimated as Dall's and harbour porpoises is based on for marine mammals in the Alaska Gyre interviews with John Ford, Kathy Heise, (Trites and Heise, this voL). Lance Barrett-Lennard, Ron Bates, and Diet compositions for resident killer Whales, Tamara Guenther. Dall's and harbour porpoises were obtained Estimates for Dall's porpoises range from Pauly et al. (1995), from Barrett­ between 150 - 300 (Lance Barrett-Lennard, Lennard et al. (1995) for transient killer Department of Zoology, UBC, pers. comm.) whales and from Olesiuk et al. (1990c) for and 2,000 - 4,000 (Tamara Guenther, harbour seals, for which the data were also Victoria Marine Mammal Research Group, broken down into estuary and non-estuary, and Dave Duffus, Geography Department of and by month, then re-aggregated. Diet the University of Victoria, pers. comm.), composition on a species basis are presented [The high estimate of T. Guenther is based in Appendix 1, Table L. on the high number of animals that wash up Table 33 summarizes our overall results. dead each year]. The estimates for harbour porpoises ranged from 50 - 150 (Lance Barrett-Lennard) to 2,000 - 4,000 (Dave Marine birds Duffus, pers. comm., with Tamara Guenther in agreement). The general consensus is that Here, marine birds are defined as birds that there are more Dall's porpoises than harbour feed extensively in the waters ofthe Strait of porpoise in the Straits of Georgia and Juan Georgia, i.e., all shorebirds except great blue de Fuca (T. Guenther, pers. comm.). This is heron have been omitted. Most of the consistent with Barrett-Lennard who also omitted birds, such as dabbling ducks,

70 Table 33 Statistics of marine mammals in the Strait ofGeorgia.

Species Mean Daily Q/B Pop. Pop. Food weight ration no biomass cons. (Units) (kg) (kg'day" ) (yea( ) (N) (t'km'~) (t'km'~'year'l) Orca (resident) female 1,974 43.28 8.0 40.2 0.011 0.09 male 2,587 53.73 7.6 40.2 0.015 0.11 Orca (transient) female 2,761 56.62 7.5 5.5 0.002 0.09 male 3,068 61.59 7.3 5.5 0.002 0.11 Dall's porpoises female 6 \.4 2.69 16.0 563 0.005 0.08 male 63.1 2.75 15.9 563 0.005 0.08 Harbour porpoises female 29.5 1.50 18.6 250 0.001 0.02 male 32.6 \.63 18.2 250 0.001 0.02 Harbour seals female 56.4 \.78 11.5 7,163 0.059 0.67 male 63.9 2.02 11.5 7,163 0.066 0.76 All mammals -- -- 11.2 16,041 0.168 \.88 geese, and swans, feed only to a very limited were taken from Vermeer and Ydenberg extent in the marine realm proper (Vermeer (1989, p. 67). The original data covered the and Ydenberg 1989, p. 62). period between May 1 and August 31 (year not specified). Mean body weight figures were obtained mostly from Palmer (1962, 1976a, 1976b, The mean body weight of Brandt's 1988) and Nilsson and Nilsson (1976), while cormorant was taken from Palmer (1962, p. population size estimates were taken mainly 345). Vermeer's estimate of the population from Vermeer (1981, p. 115) and Vermeer in March-April, 1977 was used for the and Ydenberg (1989). summer population (Vermeer, 1981, p.115). Vermeer also included a November value, Bird rations were derived from: but this was considered less relevant, since the March-April environment is more 10gR = -0.293 + 0.85 10gW similar to the summer environment. where R is the daily ration in g, and W is The mean body weight of Arctic and other body weight, also in g (Nilsson and Nilsson loons is represented by the value for Arctic 1976). The results are consistent with a loons provided by Palmer (1962, p.43), as graph in Muck and Pauly (1987, Figure 4) Arctic loons have been reported to represent illustrating the relationship between body 90% of all loons in the Strait. The weight of fish-eating Peruvian guano birds population value in this study is the mean and their daily food consumption. for March-April, 1977 (Vermeer 1981, p. 115). The mean body weight and population figures for glaucous-winged gull, double­ Western brebes represent approximately crested cormorant, and pelagic cormorant 90% of all the grebes in this study area

71 (Vermeer 1981, p. 115). Thus, the mean cormorant are from Palmer (1962, p. 349). body weight of grebes can be represented by The original data were collected from the an the average value for Western grebes west coast of Vancouver Island and Oregon (Nuechterlein and Storer (1989, p. 40). The and are not specific to the Strait ofGeorgia. population value for grebes is the mean for March-April, 1977 (Vermeer, 1981,p.115). Vermeer (1981, p. 115) reports that Arctic loons represent 90% of all the loons in the The mean body weight of surf scoters is Strait of Georgia. (Palmer (1962, p. 49) from Vermeer (1981, p. 114), as is the reported that Arctic loons feed on herring. population estimate; the latter, however, was Pacific loon stomachs obtained from Active multiplied by 0.5 to eliminate seasonal bias. Pass also contained Pacific herring The mean body weight of white-winged (Robertson 1973). scoters is from Palmer (1976b, p. 284). The population value is the average of March­ Western grebe is the most common species April, 1977 Vermeer (1981, p. 115). Other of grebe to visit the Strait (Vermeer et al. diving ducks include oldsquaws, Barrow's 1983). Robertson (1973) reported only the golden eye, greater and lesser scaup, and winter and March data, from which summer other unidentified species. Mean body food composition has been inferred. Herring weights were assumed similar to that of is an important food of Western grebes in oldsquaw (Palmer 1976a, p. 354), and values the Strait (Vermeer and Ydenberg 1989, p. for May were averaged. The population 64). estimate is the average for March-April, Surf scoters have been reported to be the 1977 (Vermeer 1981, p. 115). most common diving ducks in the Strait The mean body weight and population (Vermeer, 1981, p. 115). They feed mostly values for pigeon guillemot are from on blue mussels, but also on snails, errant Vermeer and Ydenberg (1989, p. 67). The polychaetes and barnacles (Vermeer and data were for the period between May 1 and Ydenberg 1989, p. 65). White-winged August 31 (year not specified). Other auks scoters consume mainly clams, but, like are common murre and marbled murrelet. scoters, they also feed on snails, errant The mean body weight of these auks was polychaetes, and barnacles (Vermeer and assumed similar to the value for pigeon Ydenberg 1989, p. 65). Oldsquaw eat mostly guillemot. The population is the average of bivalves and snails in the summer (Vermeer March-April, 1977 (Vermeer 1981, p. 115). and Levings 1977, p. 58). The diet composition of other diving ducks was The mean body weight of great blue herons assumed similar to that ofoldsquaw. is from Palmer (1962, p. 393). The population value is the average of the years The diet composition of pigeon guillemot 1980, 1981 and 1987 (Butler 1989, p. 114). was taken from Vermeer and Ydenberg (1989, p. 63). Pigeon guillemot feed mainly Food composition data for glaucous-winged on benthic fish. The diet composition of gull, double-crested cormorant, and pelagic other auks, such as common murre and cormorant are from Vermeer and Ydenberg marbled murrelet, was assumed similar to (1989, p. 63). The data are for the nesting that of pigeon guillemot (Vermeer and season. Food composition data for Brandt's Ydenberg 1989, p. 63).

72 Table 34 Body weights and related statistics of marine birds in the Strait of Georgia.

Species Mean Daily Q/B Pop. Pop. Food weight ration No biomass cons. (Unit) (kg) (kg'dai) (year" ) (N) (t'km'z) (t. km-z 'year"I) Glaucous-winged gull 1.16 0.30 94.49 26,000 0.0044 0.41 Db.-crested connorant 2.00 0.48 87.07 4,000 0.0012 0.10 Pelagic connorant 1.80 0.44 88.46 4,800 0.0013 0.11 Brandt's cormorants 2.43 0.56 84.59 1,661 0.0006 0.05 Loons (Arctic & other) 1.95 0.47 87.41 2,768 0.0008 0.07 Grebes (west. & other) 1.24 0.32 93.56 17,159 0.0031 0.29 Surfscoters 1.10 0.29 95.24 10,793 0.0017 0.16 White-winged scoters 1.35 0.34 92.36 1,384 0.0003 0.03 other diving ducks 0.91 0.24 98.01 24,354 0.0032 0.31 Pigeon guillemot 0.45 0.13 108.91 1,000 0.0001 0.01 other auks 0.45 0.13 108.91 3,875 0.0003 0.03 Great blue heron 2.95 0.66 82.15 812 0.0003 0.03 All birds -- -- 91.77 98,604 0.017 1.57

Food composition data for great blue Herons life history, used for grouping species were were taken from Verbeek and Butler (1989, obtained from Hart (1973). p. 75). Gunnels represent almost 50% of their food. The DFO statistical areas corresponding to the Strait of Georgia as defined by Diet compositions are presented on a per Mackinson (vide supra), are 14 to 18,28 and species basis in Appendix 1, Tables L. Table 29. Because of their importance for the 34 gives a summary of the population salmon fishery, DFO treats the lower Fraser estimates. River and estuary (areas 28 and 29, further divided as subareas 29-A to 29-F) separately from the Strait of Georgia. However, they Fisheries harvest in the Strait are included in our definition ofthe Strait of of Georgia Georgia (see Mackinson, this vol.)

(Eny A. Buchary) The same factors were used for conversion to round weight as for the catches on the B.C. shelf (see above) Data Sources and Methods Based on their life history (Hart 1973), the Fisheries catch data were extracted from the various species caught in the Strait were B.c. Commercial Catch Statistics Database combined into functional groups such as of the Department of Fisheries and Oceans demersal fish, shellfish, pelagic fish, eggs, (DFO). Data from July, August and non-food biota and zooplankton. September were chosen to represent "summer". The years covered were 1982 to The demersal fish box includes halibut, 1989 (no computerized data were available lingcod, Pacific (grey) cod, sablefish (i.e., for the years prior to 1982). Information on black cod), brill sole, Dover sole, lemon

73 2 Table 35 Commercial fisheries harvest (kg·km- • Another component is "eggs" year-I), Strait of Georgia. (Summer values only; which includes both salmon roe and based on DFO data for the years 1982 to 1989). herring roe, while "non-food biota" consist oforganisms used for "mink Functional Groups Areas Areas Whole 28&29 14 - 18 Strait feed, scrap and reduction"; the Bivalves 7.1 405.04 412.2 "zooplankton" group consists solely Sea cucumber/urchins 4.02 25.6 29.6 ofeuphausiids. Octopus 0.07 1.47 1.54 Squids 0.006 0.09 0.1 Table 35 presents estimates of Plankton 0.9 0 0.9 summer catches for these and other Codslblack cods 6.9 22.95 29.8 functional groups, by sets of Prawns/shrimpsicrabs 176.2 48.1 224.3 subareas ofthe Strait ofGeorgia. Large pelagics 0.02 13.2 13.2 Small pelagics 0.7 18.9 19.6 Misc. demersals 3.6 6.6 10.2 Ocean perches/rockfishes 2.8 19.8 22.5 Balancing the Strait of Sharks 0.95 34.5 35.4 Georgia Model Transient salmon 2065.1 327.4 2392.5 Resident salmon 99.4 176.8 276.2 Others 0.2 0.04 0.2 (Judson Venier) Sum 2367.9 1100.34 3468.24 Ten functional groups ("boxes") sole, rock sole, Pacific Ocean perch, were originally identified for greenies (i.e., yellow tail), rockfish, red inclusion in the Strait of Georgia model, but snapper (i.e., yellow eye), hagfish, flounder, problems arose during the process of skate, (silver perch), turbot, pollock and parameter estimation and more boxes had to hake. be incorporated. For example, demersal fish were split into two boxes (hake and DFO use "shellfish" as a catchall term for all demersal fish) because of excessive invertebrates. Major components include cannibalism within the original box. After a benthic organisms such as red sea urchin, meeting with the parties involved in input prawn, shrimp, razor clam, butter clam, estimation, initial guesses of B, PIB, Q/B, Manila clam, native littleneck clam, catches and of diet compositions (Des) were Dungeness crab, red rock crab, geoduck identified which were questionable, and thus (Harbo et al. 1992), horse clam, scallop, provided leeway for balancing the model. crayfish and sea cucumber, but also includes The changes thus effected were as follows: non-benthic organisms such as squid.

Salmonids, (recorded as red and white Mammals and Birds chinook, sockeye, coho, pink, chum and steelhead) are the dominant pelagic species. The top predators (birds, resident mammals, Aside from salmonids, other species, i.e. transient orcas) were believed to have the eulachon, sturgeon, tuna, smelt, dogfish, best data and few of the inputs (in Wada, shark and herring also contribute to the this vol.) needed to be modified. The PIB pelagic box. (See also Buckworth, this voL). value of resident mammals was too low to accommodate the predation pressure exerted

74 by the transient killer whales and was Small Pelagics increased by 0.2 year"\. We believe this change to be justified because the bulk of P/B was assumed to be higher (by 1 year"!) the biomass ofresident mammals is made up than the value for Large Pelagics. The Q/B of harbor seals, which have a fairly high value was set at a high 18 year" \, meant to reproductive rate, and hence a high P/B reflect the high feeding activity of small ratio. pelagic fish. An EE value of 0.95 was input to obtain a reasonable biomass estimate, and Also, part ofthe DC's ofResident Mammals to reflect our assumption that most of the and Birds was changed to reduce predation small fish production is consumed within the pressure on Miscellaneous Demersals. ecosystem. Resident Mammals were changed to feed more heavily on salmon and Birds were changed to feed more strongly on krill Hake (Carnivorous Zooplankton). The biomass estimate inrut into the model was increased by It·km" over its value in Salmon Venier and Kelson (this vol.), and the PIB value slightly increased, from 0.72 year"! to Salmon were incorporated as an individual 1 year" \, to accommodate the high predation box. In this model, salmon catch was 0.743 pressure to which this group is subjected. t'km"2year"\, B was 3.8 t'km,2, their PIB was The Q/B value was reduced by 0.84 year" \ 0.75 year"!, and Q/B was 0, i.e., they do not from that in Pitcher (this vol.) to decrease act as predators in the system. Large pelagic the predation by hake on the Miscellaneous harvest was lowered to include only dogfish. Demersals, whose biomass would otherwise The outputs were generally similar, but EE shoot up. The DC was shifted to increase of large pelagic fish was reduced by -0.5 predation on krill and decrease that on because of decreased fishing pressure and a Miscellaneous Demersals, for the same shift ofpredation by resident mammals from reason. large pelagics to salmon. This model is more representative of the system, as it includes the B of salmon and the influence, although Miscellaneous Demersals minimal, that they exert on the other The biomass estimate of 13 t'km-2 used in components. the model is more than twice as large as the value in Venier and Kelson (this vol.). Large Pelagics However, the original estimate was based on the large, commercially important species Biomass was estimated to be -15t·km,2 for which catch statistics are available, and which is 1.5 t·km·2 lower than the estimate did not include the many smaller species in Buckworth (this vol.). P/B was reduced also found in the ecosystem. Thus the from 3.5 year"\ to I year" \, the latter being a increase seems justified. For the same more realistic value for larger fish. DC was reason, the P/B value used here is 1 year"l, adjusted to reduce predation on which is higher than the mean in Venier and Miscellaneous Demersals and increase that Kelson (this vol.). The EE value estimated on krill. from the model is very high, indicating that

75 this group is heavily fed upon within the estimated from the literature is probably too system. low for such small organisms.

Jellies Carnivorous Zooplankton The :high biomass estimate of 312 t·km-2 in Venier and Octeau (this vol.) pertained to an The biomass estimate was obtained from the annual maximum from an specific area of model by entering an EE value of95%. This the Strait, which cannot be considered value is reasonable as most ofthe production representative of the entire system and from this group can be assumed to be period considered here. The abundance of consumed within the system. PIB is higher gelatinous organisms within the Strait is than the maximum estimated in Venier and very habitat specific and numbers can be Octeau (this vol.) but lower than that very high on a local scale within some ofthe estimated for other models (e.g. Christensen fjords and bays. Here, the biomass for the 1995). The QIB value was raised to yield a 2 Strait as a whole was set at 15t'km- , positive estimate ofrespiration. pending a comprehensive study. The original values of P/B and Q/B values in Herbivorous Zooplankton Venier and Octeau (this vol.) were also adjusted downward, another reflection of The biomass used for this group is twice that our difficulties with the conversion of site­ one estimated in Venier and Octeau (this specific, dry weight-based estimates. vol.); similarly, a P/B value of 55 year-I was used instead of the value in Table 24, to Large Macrobenthos accommodate predation. Christensen (1995) used a similar value for planktonic groups in The biomass value of 140 t·km-2 is based on the North Sea ecosystem. The value of Q/B the work of Ellis (1967-1968), cited by in Table 24 was raised for the same reason Guenette (this vol.) A P/B value of 2 year-! as for Carnivorous Zooplankton. These was used, which is slightly less than that changes imply that the original parameters estimated value for carnivorous may not have been realistic. macrobenthos (in Guenette, this vol.) because the group contains omnivores as Primary Producers well. The Q/B value in Table 27 was lowered slightly to reduce predation on Biomass was estimated by the model after Small Macrobenthos. entering an EE value of0.40 and a PIB value of 125 year-I. This EE value reflects our Small Macrobenthos assumption that the bulk of primary production is not directly consumed within The original B estimate of 375t/km2 in the system, but instead cycled to the detritus. Guenette (this vol.) had to be increased to The P/B value is intermediate between an 400 t·km-2 to accommodate predation. P/B upper estimate, pertaining to phytoplankton also needed to be raised for the same reason and a lower estimate, pertaining to the and also because the P/B value of 0.6 yea{l macroalgae.

76 Detritus During the summer, large increases in zoo­ plankton biomass arise following the large The biomass of this group was estimated blooms of phytoplankton in the late spring. using the empirical formula: . This increased production of plankton in early summer supports the system loglOD = -2.41 + 0.954 loglOPP + 0.863 loglOE throughout the rest ofthe summer months. where D is the detritus standing stock in 2 The Strait of Georgia ecosystem is a gC'm- , PP is the primary production in 2 l complex one with many trophic energy gC'm- 'year- , and E is the euphotic depth in flows (Figure 4). Many components in the m (Pauly et al. 1993b). ecosystem appear to be generalists, feeding on a high diversity ofprey. For example, the Results and Discussion resident mammals' omnivory index (the heterogeneity ofthe DC values) is extremely The basic parameters for the strait high. They feed at several separate trophic ecosystem are presented in Table 36, while levels, and their own trophic level is Appendix Table 0 shows the corresponding comparatively low compared with that of diet matrix. some of their prey items. One of their prey species, hake, indeed has a trophic level As can be seen, both zooplankton types, higher than theirs. The varied nature of small macrobenthos, hake, and small many groups' diets and the high variety of demersal fish were all consumed heavily in trophic levels present in the system may the system. Large pelagic fish, salmon, makes successful single-species manage­ jellies and detritus were not used much as ment of the specific components of the sources of food: their EE values are low. ecosystem rather difficult given their interactions (Figure 5). Thus the Table 36. Basic parameters for the Strait of Georgia 2 1 use of ecosystem models, such as model. Biomasses are given in t·km- ·year- , while production/biomass (P/B) and consumption/biomass Ecopath, as tools in the formation (Q/B) both are annual rates. EE is the ecotrophic of management decisions are efficiency expressing the proportion of the important for proper production that is lost to exports or predation management. The more one mortality. understands the system as a whole, the more one can under­ Group Biomass P/B Q/B EE stand its components, and better Mammals (Res.) 0.153 0.400 11.540 0.605 Lg. pelagics 15.000 1.000 5.000 0.821 management decisions can result Sm. pelagics 24.160 2.000 18.000 0.950 from these ecosystem analyses. Hake 24.500 0.900 5.000 0.879 Misc. demersals 13.000 1.000 4.240 0.985 The food consumption of birds is Jellies 15.000 3.000 12.000 0.137 very high, due to their fast Lg. macrobenthos 140.000 2.000 8.750 0.260 metabolism. They consume as Sm. macrobenthos 400.000 3.500 23.000 0.631 much food as both groups of Cam. zooplankton 16.577 12.000 40.000 0.950 Herb. zooplankton 16.269 55.000 183.333 0.950 marine mammals combined even Prim. producers 250.100 15.000 0.000 0.760 though their biomass is much Birds 0.017 0.100 91.770 0.000 smaller. Marine birds are also Trans. Orcas 0.005 0.200 7.400 0.000 feeding at a wide range of trophic Salmon 3.800 0.750 0.000 0.403 Detritus 7.000 -- 0.967 level. 77 --. Flow --' Comeclor --* Harvellt --+ Flow to detritu. l'ran.ient Orcu -. Respiration ...r- Import T .--- I 4 r- Mammal (Res.) Hake 76.0 ~27.2 rae pelagic r. Mi,c. dcmersals 0.4 0.0 n· I. 1-+ 0.0 17.7 8.6 I 4S.0 11.2 G) Birds ~ > I 0.1 G) 1.2 0.3 -....J .J Jellie• s~all pclasiu +1 74.8 00 () 3 0.0 k. .- 99.0 or- SA s:. 299.6 89.4 Q, 0 '" III .. --··i !---- H Cam. ; 00 I. .\.- 17S 7 1 4S2 2 Lg.Macrobenthos p02 . I--- . 298.4 0.0 iii 700.0 61.3 +- Sm.Maerobenthos I---- 2 o4 Herb. looPI.] 23S7.2 b . .f1 8 .. 460.0 894.8 S960.0 I T

I - Salmon Detritus rrim. Produce" L +- 0.7 1.7 901.2 SOOO.O

Figure 4 Flow chart of trophic interactions in the Strait of Georgia. All flows are expressed in t.km-2'year-1 and all biomasses in 2 t·km- ; minor flows are omitted, as are all backflows to the detritus. _ .- 8 S ~ .~.Il ft 1 ~ i ~ ~ ~ t ~ . 1 i t II 1 £ • il !! .j ~ ~ t ~ ~ ~ J ~ ~ ~ i ~ ~ 1 '5 ::E j ~ ~ i ~ ~ ~ ~ ~ ~ ~ ~ ~ o ii: Mammal, (Res.) ------'---- l..4rse pelagic,

Small pelagics

lIake ... ~------;::) 0 Misc. demersai. 0:: 0 0 Jollie. :z t- I.g.Macrobelllho. U <: c.. ::E Sm.MaclObcnlhns - imllmilI lillII'mJ .....:I \C) Cam. Zoopl.

Ilcrb. Znopl. lBlfIIIl~

Prim. Produce.. ~

Birds Transient Orcas. Slllm(N\

l),;lrilu.

Fishel}' ~

Figure 5 Mixed trophic impacts in the Strait of Georgia ecosystem. A few problems were identified in the localized fauna. Integrating data boxes was balancing of the model which warrant often difficult because many estimates were discussion. Migrations into the strait of large obtained from specific sites with species not quantities ofsalmon had to be accounted for. found in other sites. Therefore, finding These fish enter the strait in mid to late common input values for aggregated groups summer, spend little time there, then enter of organisms that could be used to run the the rivers to spawn and die. They are model was a difficult task (this was consumed by resident mammals, especially particularly evident for the zooplankton box, orcas, and in turn, consume almost nothing whose parametrization appears quite while migrating. They contribute to the problematic). This resulted in the drastic detritus as their carcasses are carried down modifications of some P/B and Q/B values the Fraser and other rivers and into the Strait when the model was balanced. of Georgia. Although they do not affect the ecosystem as a whole very significantly, they are nevertheless one of its components. Conclusion The resident mammals depend on them for a food source and there are large fisheries for Through this run of Ecopath, a clearer them throughout the Strait. Their picture of the highly complex ecosystem of contribution to the detritus, although small the Strait of Georgia was presented. The in comparison with other components of the steady-state representation of a naturally system, is an important source of nutrient dynamic system and the identification of influx in late summer when other nutrients shortcomings in the data should provide have been depleted. good reference points for future studies concerning the strait. Models such as Further studies need to be conducted in Ecopath have implications in the proper order to get a clearer picture of the management of resources which are intricate ecosystem functions in the Strait. A problem components ofthe ecosystems in which they when collecting data for input into the live. Understanding the processes and model was their heterogeneity. The Strait of interactions within the systems, and that Georgia has a very varied coastline impacting one group will cause changes in consisting of bays, inlets, fjords, muddy another, can aid in the formation of long­ shores, estuaries, rocky shores, as well as a term management policy. vast open pelagic zone, each harboring

80 GENERAL DISCUSSION Power

The real power ofa model is not so much in the result, as in the process, in the learning A Road Test of Ecopath that it engenders - it's not your destination that matters, but the fun you have on the (Rik Buckworth) way there! Students reported that learning to drive Ecopath forced them to think about a A road test, you might ask? Why report the whole system, emphasizing both the amount construction of a mass balance model as "a ofappropriate information available, and the road test"? A new car model is reported to data gaps. As this information - biomass, the motoring public on the basis of some diet composition, etc., had not previously characteristics with which they are been combined in such away, Ecopath reasonably in tune, usually by a motoring allows users to evaluate the importance of writer who is not otherwise familiar with the gaps. The system thus has two important vehicle - and need not be an especially good attributes ofa useful model - it predicts, and driver - but who has a wide range of it has the capacity to surprise, as shown by experience with different cars. This note the total fish consumption by mammals and uses the analogy of a road test as a birds in Georgia Strait (see Wada, this vol.). framework for reporting the experience of a group of graduate students in using the Ecopath system to model the Strait of Handling Georgia ecosystem. An attractive, almost seductive, attribute of Ecopath is its simple data requirements. The scene ofthe road test Estimates of biomasses, consumption rates, diet, etc. are often easy to come by. The Strait of Georgia separates southern However, it is important to emphasize that British Columbia from Vancouver Island. It the requirements are not simplistic. Immense is a long, narrow and relatively sheltered amounts of work may be represented by a basin with a total area of around 7000 km2 few numbers, and conversion of information and mean depth of 156 m. Constrictions at between different units or currencies can each end by narrow straits, a fjord coast and make the derivation of those numbers the presence of several islands create strong problematic. But there is a danger, too. Sins tidal flows. The environment of the southern of omission can occur through plain half is strongly influenced by the Fraser ignorance: an important component of an River (Mackinson, this vol.). The Strait and ecosystem may be unrepresented in the its tributaries thus provide a diversity of literature, or lack of knowledge about habitats, including those for several functioning of an ecosystem (such as the commercial and recreational fish species, timing and degree ofmovements ofanimals) including Pacific salmon, as well as a suite may mean that a model represents reality of marine mammals and birds (Wada, this very poorly. vol.).

81 Fuel economy and Price Suggested Improvements for Ecopath Modeling At first glance, Ecopath can appear deceptively fuel hungry. It requires simple data inputs that may really have been (Carl Walters) rendered down from a vast amount of information - there can be lots of legwork needed, lots of searching for information Moving to Dynamic Predictions Using that just may not be there. We estimated that Mass Action (Volterra) Assumptioni the development of inputs for the Strait of Georgia model cost around 300 person Ecopath could be extended to either (1 ) hours. Balancing the model and reporting estimate the rates ofeffective search aij from can probably be accomplished in something the flow and biomass estimates less. But is this really expensive? (aij=flow/[BiBjD, or (2) constrain and possibly improve the flow estimates given Fuel quality is important to Ecopath: it will independent estimates of aij. Given the aij go further if good fuel is available. Expertise estimates, Ecopath could be used to make on trophic groups in the area being modeled transient, non-equilibrium predictions for at substantially reduces the amount of legwork. least modest perturbations in the flow rate It thus took a mere 200 person hours to structure (all that is needed here is a simple assemble the information for the more differential equation solver). complex Alaska Gyre model. For non-equilibrium modeling, trophic flows Market between the biomass pools representing consumers and their food have generally Who might use Eopath? I see fisheries been modeled as the product of a functional applications of the system as a means of response representing the consumption rate testing understanding of, for example, the per consumer, times the consumer biomass. bounds on biomass or exploitation rates of a fished species, or generating hypotheses floWij=f(Bi)Bj where about the impact of fishing on different i = resource species, j = consumer ecosystem components. But the application should be broader - wherever there is need to Usually we use a nonlinear, saturating understand a system or guess at its behavior, functional response ofthe form Ecopath has potential application. f(B)=aBi/(l +aBi/cmax) where Cmax = maximum consumption per consumer. Conclusions But when we run such models, and when Yes, the Ecopath system is powerful - but rates of effective search "a" (volume per drive carefully! time) are calculated from field data on

I Since the workshop was held, a dynamic simulation model (Ecosim for Ecopath), including the ideas presented below, has been developed and described (Walters et aI., in press). 82 search efficiencies, and when stomach of reacting to prey that enter the swept contents ofconsumers are examined to see if volume. For zooplankton, there are lots of consumption rates are near Cmax , it is usually aij estimates (called "filtering rates" in the found that feB) is small compared to cmax ' literature). There are some estimates for That is, consumers in nature most often act planktivores, and a few for piscivores. as though f(B)=aB, with only modest effects from handling times/satiation. There is in fact an evolutionary argument due to Labeling the probability distributions for Crowley (1975) that "optimal" consumption output (derived) variablei rate parameters in nature should generally reduce "a" enough to make consumption Ecopath computes values of some "output" flowlbiomass variables needed to provide rates be well below Cmax (working hard mass balance (equilibrium in flows and enough at feeding to approach Cmax is often not a wise evolutionary strategy). biomasses) given "input" values for other variables. The Ecoranger routine Thus data and evolutionary arguments (Christensen and Pauly 1995) allows support the use of simplified mass action Ecopath users to specify probability (Volterra) models for the flows, ofthe form: distributions for the input variables, and uses a Monte Carlo procedure to sample from flow"=a"B'B'IJ IJ 1 J these input distributions so as to generate probability distributions of the output For independent estimation of the aij it is variables. Such Monte Carlo simulation useful to note that the units of these procedures are perfectly acceptable (and parameters are area or volume "swept" for indeed probably most efficient) as a method prey per unit of time per unit of consumer for numerical assessment of probability dis­ biomass (units are arealtime biomass in tributions that are complicated functions of aerial units, volume/time for biomass in per probabilistic inputs. During the workshop, a volume units). Variation in aij among prey question has arisen about whether the types i represents "preference" by consumer Ecoranger output distributions can properly j created by either variation among prey in be labeled "Bayes posterior" distributions. the consumer's willingness to take them Strictly speaking, the Bayes posterior when it encounters them, or variation among distribution for a parameter is the prey in habits/behaviors/morphological conditional distribution for the parameter characteristics that affect vulnerability to the given the data, and thus, the first answer to consumer. this question is no. The Bayes posterior The simplest possible estimate of aij is probability assigned to parameter value b given a data set Y is the ratio pCb IY) = reaction field width (or area for volumetric p(Y Ib)Po(b)/p(Y), where Po is a prior prey biomass units) times the sum of probability assigned to b and p(Y) is the average swimming speeds of the predator total probability of obtaining Y (integrated plus the prey, divided by the average weight of a predator individual. Generally the a's are much smaller than this estimate would 2 Since the workshop was held, Ecopath was modified to incorporate the semi-Bayesian approach indicate, due to relatively low probabilities described below.

83 over all possible values of the parameters, This is a bit more than haggling over i.e. the integral of p(Y Ib) Po(b) over all b terminology. Suppose a prior distribution is values). If one thinks of the Ecopath inputs provided for the b, poCb), which is then as Y and the Ecoranger probability compared to the "tighter" distribution p(b) specifications for Y as p(Y Ib), there is an that is produced by Monte Carlo routine in apparent resemblance to the Bayes posterior Ecoranger. Clearly, information about b has distribution since the Ecoranger simulation been gained. But this information came not procedure also maps pCb IY) from p(Y Ib). from gathering "data" in the classical But the mapping is not as simple as Bayes statistical sense, but rather from the formula above. structural relationship b = fey), defined by the Ecopath model. Combining this The Ecopath model structure provides a structural information with prior unique value ofb for each input combination probabilities p(Y) to generate p(b) is similar Y that is biologically (and to combining data with priors to generate thermodynamically) feasible, and also posteriors in Bayesian statistics. But there defines combinations Y that would imply are those who would argue that combining biologically impossible b values. Thus, what structural knowledge with prior probabilities Ecoranger does is to generate the probability is not the same as combining data with distribution for a transformation or function pnors. (b) of the input variables Y. That is, it is finding the marginal distributions p(b) of To be safe, and to avoid silly arguments functions beY), given the distributions p(Y) about Bayesian analysis and what data from which the functions bey) are "really" are, Bayesian terminology should calculated. This is a perfectly valid be avoided unless one is indeed predicting b probabilistic calculation [the textbook case values for which there are (1) particular is: "find p(u) given p(y) and assuming u = observed values, and (2) likelihood f(y)"] whether or not a Bayesian functions for these observed values, derived interpretation is given to the p(b) by viewing by analysis of the sampling process leading the p(Y) as "priors". If the p(Y) are viewed to the observed values. At least some as Bayes prior credibility measures, then Ecopath assessments and their strictly speaking the p(b) distributions corresponding field estimates may be treated generated by Ecoranger should be termed this way in the future. Here, it is necessary "derived Bayes Prior Distributions for to be careful about the form and parameters output parameters b". In this phrase, of the distributions specified in Ecoranger "derived" means that p(b) is derived from for such b values, since these distributions p(Y) by using the fact that each Y value have to be regarded as likelihood functions implies a unique b value, which has a non­ p(bmeasured Ibtrue) rather than as specifications zero probability only if it is biologically and ofprior belief about the true value. thermodynamically feasible. [Among other things, all Ecopath "boxes" in an model Another development that would move must have parameter values within Ecopath more directly into the traditional acceptable ranges, and their consumers Bayesian approach would be the inclusion in cannot consume more than the boxes Ecopath models of derived quantities not produce.] required for the mass balance solutions, but

84 for which there are observed data or (computed from Y variables), and that the "likelihood" expectations. For example, product of these p(b) values over b were Ecopath could be used to predict rates of treated as though it were the likelihood Lk- effective search for predators j on prey types Then, when the (Y, b) subsample from the i, as aij=flowijlBiBj- If likelihood functions resampling process is examined, it will are specified for the observed input values of generally be found that the sample such derived quantities, then the product of distributions p(Y) for the inputs is different such likelihood, for any specific Ecopath than originally specified; that is, including input parameter combination Y, is an p(b) information in the resampling process additional measure of the credibility of that leads to the impression that one has gained combination Y (beyond the p(Y) measures some information about the inputs Y as well included using Ecoranger). as the outputs b. This information arises not from data in the traditional Bayesian sense, To include likelihood functions for some b but rather from applying the relations among and derived quantities in the Ecoranger Y anif b implied by the Ecopath functional calculations, the Monte Carlo sampling structure. So what is being done, technically, procedure must be modified slightly, so as to is finding the joint distribution for (Y, b) perform "Sampling/Importance Resampling" implied by the marginal priors p(Y) and p(b) (SIR, see McAllister et al. 1994). This under the relational constraints on (Y,b) procedure is very simple and consists of the implied by Ecopath structure. I do not know following steps: do the Monte Carlo of a standard technical term to describe this simulations as at present, and for each ofthe joint distribution, but it could perhaps be k= 1...n sampled parameter combinations called a "joint prior" or just the 'joint (Y,b)k, store the (Y,b)k value and the distribution given Ecopath relations and likelihood function value (product of constraints". When the sample frequency probabilities of observed values b, aij, etc., distributions for individual Y and b given Y) Lk for the sampled combination. parameters selected by the resampling Then, instead of summarizing and plotting process are plotted, the resulting individual the results ofthese n sample trials, plot only frequency patterns could be referred to as the results from a subsample m of the trials, the "marginal distribution under Ecopath". where the probability ofeach sample k being However, each distribution should only be included in this subsample (resample) is just referred to as a "Bayes marginal posterior the sample "importance weight" wk, i.e., distribution" if the p(b) has been structured explicitly as a likelihood function simply Lk divided by the sum of the Lk p(bobserved Ib) for some particular observed values over all n samples. There are more values(s). elaborate procedures for obtaining the wk weights, but these only apply when the Y input samples are chosen from distributions Incorporating Seasonality or criteria other than the Ecoranger priors for these Y. Ecopath does not actually require equilibrium on seasonal time scales. Its Suppose a SIR algorithm is being set up balance relationship is on total flows over a wherein prior distributions p(b) have been defined time step. Biomasses (B) are specified for the Ecopath derived variables assumed to be the same at the end as the

85 start of this step, but may vary arbitrarily 2. Provide a graphical interface for the user over the step. However, to correctly specify to shape the Bi/BMAXit, and other ratios P/B and Q/B ratios, and consumption listed above in, say, monthly steps; patterns, at least the relative pattern of B change in time (t) and these ratios must be 3. On exit from this seasonal interface, set defined by the user. the model G·P/B j weights by summing over the n seasonal values (with n = 12

Consider production Pi = (P/B)i B j for monthly values) ofthe ratios;

The annual total can be calculated as a sum 4. Feed the known B, P/B, etc. values into over t=1,...., n time steps within the year as Ecopath as usual and rescale output calculations using the G functions as appropriate to provide both seasonal maxima and annual averages. If we let P/BMAXi be the maximum value of P/B it over t, and BMAXi be the maxlfnum BMAXi over t, then the above sum can be Exploring ecosystem responses to written as environmental variation3

(J.J. Polovina) or Considerable progress has been made in modelling links between ocean physics and plankton dynamics on seasonal, inter­ where R-P/B = (P/Bit)/(P/BMAXi) and (R·B ) it it annual, and decadal scales (Fasham 1995; = (Bit)/(BMAXi) are relative P/B ratios and Polovina et al. 1995; Sarmiento et al. 1993). biomasses by season, respectively, and However few studies have attempted to GP/B is the sum of products of the relative i model impacts to higher trophic level. While P/B ratios times relative biomass. physical parameters are not explicitly To use this approach, a way must exist for formulated in Ecopath and Ecosim, these the user to specify seasonal patterns in the ecosystem models can be used to explore ecosystem responses to physical variation. relative biomasses, RBit = (Bit/BMAXi), products (RP/B ) = (P/Bit)/(P/BMAXi), it Ecopath provides a tool to examine consumptions (Q/Bit)/(Q/BMAXi), and in the ecosystem structure and function in different diet composition matrix DC . ijt climatic states. For example, it now appears Operationally, this reduces to just four that both the physical and biological regimes changes in Ecopath: in the Gulf of Alaska gyre were different between 1960-76 and 1977-88. In particular, 1. Warn the user to input seasonal maxima for known quantities, rather than averages; 3 This section was written in October 1996, and its references to "Ecosim" pertain to the software implementing the dynamic simulation approach suggested above by Walters (this vol.) and fully documented in Walters et al. (in press).

86 zooplankton and salmon abundance were In many cases the functional relationship substantially greater in the Gulf of Alaska in between a physical and biological variables 1977-88 than in 1960-76. An Ecopath model are not known or the physical time series are constructed for each period could be used to not available. Ecosim can still be a very evaluate whether the increase observed in useful tool to explore ecosystem responses. salmon is simply a bottom-up response of The approach in this case would be to drive the observed increase in zooplankton, or the model directly with time series of whether more complex trophic changes were Ecosim biological variables such as involved. biomass, growth, mortality, etc., which represent the response to a physical change. Ecosim provides a tool to examine the For example, in the Northwestern Hawaiian dynamics response of ecosystems to Islands, the carrying capacity of all trophic environmental changes. One approach is to levels appears to have dropped by about explicitly describe biological parameters 50% over a 10 year period. One hypothesis currently in Ecosim as functions of physical is that this was a response to a 50% drop in variables and then drive Ecosim with primary productivity. To explore this physical time series. For example, since hypothesis, Ecosim could be driven over a plankton dynamics can be describe by a 10 year period with declining primary system of differential equation driven by production to evaluate how the ecosystem mixed layer depth, temperature, light, and responds to a 50% drop in primary nutrients, these equations could be production over 10 years. A second incorporated directly into Ecosim to couple hypothesis is that this decline was due to plankton dynamics with higher trophic increased larval advection. The ecosystem dynamics. While this will result in an impact of this hypothesis could be evaluated Ecosim model with physically driven by an Ecosim run simulating a 10 year plankton dynamics, modelling plankton period with increasing mortality (or dynamics with weekly temporal resolution decreasing recruitment) of all species groups as is done with these differential equations in proportion to the time the spend as may not be particularly useful when we are passive larvae. interested in ecosystem dynamics on scales of years and decades. However, this same Thus in summary just as Ecopath and approach may prove more useful at higher Ecosim can be used to explore impacts of trophic levels. For example, sardine survival exploitation strategies, they can also be or recruitment appears to be linked to wind used, in very similar fashion, to explore speed following a dome-shaped relationship impacts due to environmental variation. I am (Cury and Roy 1989). This relationship optimistic these tools will lead to new could be used to modify sardine survival or insights in ecosystem dynamics and at the biomass in Ecosim and then Ecosim could very least help us to get beyond the often be driven with a time series ofwind speed to encountered dichotomy that all ecosystem examine the response of an upwelling change is either entirely due to environment ecosystem to inter-annual changes in wind variation or entirely the result offishing. intensity.

87 Integration of local environmental potential target species: their preys and other knowledge4 associated species, and the weather, current, tide, phase of the moon, to name but a few. They will also compare and balance their (Nigel Haggan) observations on this particular fishing day with related experience on other days in the There has not so far been a great deal of same season. To this, they will add records crossover between Local Environmental and recollections of previous years and the Knowledge (LEK) and scientific knowledge. information which has been handed down to This is due not so much to a lack of respect them. Similarly, when Ecopath models are for LEK, after all, a great many scientists constructed by a group of experts (or based rely on it during their research. The problem on published expert knowledge handed is more one of format. Science is precise down through the scientific literature), a but, partly as a result of its precision, its number of interrelated factors must also be language is generally impenetrable at the accommodated simultaneously, and community level. With notable exceptions rendered mutually compatible. such as Johannes (1981), scientists have found LEK to be too diffuse, uncertain and This similarity may be deep enough for difficult to replicate. Also, with the Ecopath models to provide a framework for exception of some countries in the South integrating LEK, and thus strike a chord at Pacific, few attempts have been made to the local community level. Indeed, this integrate LEK into fisheries management integration may lead to cross-validation as in systems. Johannes (1981), where LEK and scientific knowledge about the target species of How can Ecopath, a mathematical model Palauan fishers where found to be mutually living in computer circuits and feeding on compatible, and where incompatibilities led quantitative data help? Would it have served to new insights, sometimes for the fishers, to have fishing community and Aboriginal but often for the scientists as well. Two community members at the November approaches are suggested to test this Ecopath workshop? Probably not for the proposition. reasons of format outlined above. Yet, there are parallels between Ecopath and LEK, The first approach is to incorporate as much notably that they both are more concerned LEK as possible into the existing databases with relationships, interactions and used as data sources by Ecopath model connections within an ecosystem than with builders, notably FishBase (Froese and achieving a deep understanding of the Pauly 1996). This relational database, now isolated elements. In their own way, both available in form ofa CD-ROM (see McCall Ecopath and LEK are intuitive. To be and May 1995), covers the fishes of the successful, fishers exploiting the resources Northeastern Pacific rather well, and its of a given area must consider an entire 1997 release will include detailed constellation of factors along with the information on the fishes of British Columbia. This database can accommodate local knowledge that is species-specific (e.g. 4 I am grateful to Rosemary Ommer for this term, which includes both indigenous knowledge and the that in Compton et al. (1994), pertaining to knowledge ofcontemporary fishing communities. the role of Catastomus macrocheilus in the

88 Secwepemc/Shuswap culture), through a correlating local knowledge and intuitions series of interrelated tables (Pauly et al with the relationships, data gaps and 1993a, Palomares et al. 1993), thus enabling conflicts identified through the model. comparisons between LEK and scientific Without wanting to prejudge the results of knowledge, and cross validation. Field and such an exercise, there is little doubt that literature-based projects devoted to such cross validatioll would lead to new accumulating and encoding such knowledge, insights and directions for future e.g., in the conteXi ,)f Cury's (1994) theory collaboration to addre~s, and perhaps resolve of site fidelity in spawning fishes, would some data gaps and conflicts. seem particularly worthwhile, given that FishBase, once it incorporate this knowledge, could be used to disseminate it Updates on Ecopath development in schools" comrnurtity centers, etc. and applications The second approaGh would consist of a workshop that would be convened to present (Daniel Pauly and Villy Christensen) an Ecopath model of a coastal area to a group of knowledgeable fishers from First The workshop repOlted upon in this report Nations and the cummunity-at-Iarge. This did not only gem'rate three models of workshop would begin the process of ecosystems important to the fisheries of

I €/I = 1l.km'.yea('

Alaska gyre

Southern B.C. shelf

Strait of Georgia

Figure 6 Trophic pyramids representing the three ecosystem models documented here. The pyramids are scaled such that the volume at each (trophic) level corresponds to the sum of all flows at that level, while the topangle is invenely related to the transfer efficiency prevailing in the system (acute angle = high efficiency). These pyramids allow direct comparisons ofwhole ecosystem properties.

89 British Columbia (see Figure 6) - though before the SIR approach was incorporated this, by itself would have been enough to into Ecopath, and hence the power of this meet our stated goals. Rather, the workshop approach will have to be demonstrated became a watershed in the development of elsewhere (e.g. with the next iterations of the Ecopath approach and software-and these models). probably for the discipline of ecosystem modeling as a whole-thanks to the input of We have developed a hybrid design for the Carl Walters (see above), who contributed incorporation of item (2) in Ecopath, i.e. for to: explicit consideration of seasonality, which we envision as driven by seasonally-varying 1. re-expression of the Monte-Carlo routine empirical observations of biomass, catches, built into the ~-release of Ecopath 3.0 and diet compositions, and using a monthly (Christensen and Pauly 1995) in form ofa cycle of temperature to force changes in P/B semi-Bayesian approach for considering and Q/B, both known to be temperature­ uncertainties; dependent (Pauly 1980, 1989).

2. explicit incorporation of seasonal cycles This design is to be implemented later in of biomass and related parameters when 1997, and hence Ecopath users wanting to mass-balancing a model; and explicitly account for seasonality will continue, for a short while, to have to 3. reexpression of the system of linear construct one model per month or season of equations behind Ecopath (see Box 1) a seasonal cycle, as in Jarre and Pauly into a system ofdifferential equations that (1993), and in Jarre-Teichmann (1995, can be integrated over time, and hence respectively. The three models presented generate fully operational simulation above all represent "summer" situations and models from Ecopath files. colleagues interested in the ecosystems these models represent should complement these Ecopath 3.0 as now distributed (free of with "winter" models, for which we present charge, see Christensen and Pauly 1996) much of what is required in terms of incorporates item (1), [in the semi-Bayesian biomasses or population sizes, catches, diet context proposed by Walters (this vol.), i.e., compositions, etc. the SIR approach of McAllister et al. (1994)], and hence can be used to generate, Then, later the summer and winter models besides distributions of the output values, can provide the input for "annual" models, posterior distributions of the input values, explicitly taking seasonal changes of given (uniform, triangular, or normal) biomass and fluxes when establishing mass distributions ofthe inputs. balance.

This we hope will go a long way towards Item (3), the use of Ecopath assessments to overcoming the doubts of those who feel parametrize the system of differential that mass-balance models - or ecosystem equations required for simulation models models in general - do not sufficiently take has been fully conceptualized (Walters et al. account of uncertainties. However, the in press). Also, a software module (Ecosim) model in this report has been constructed has been developed whose stand-alone

90 version (which reads unmodified Ecopath However, these earlier application files) is presently being tested at the successful as they were in integrating a vast Fisheries Centre, while also being modified amount of data, and describing the such that it can incorporated as a routine of functioning of these ecosystems, remained ­ the next (4.0) release of Ecopath for some colleagues at least - under the dark (Christensen and Pauly, in press). clouds of assumptions ofdeterminism (there were no procedure to explicitly account for Ecopath has been used so far to describe uncertainties) and equilibrium (change was nearly hundred aquatic ecosystems as well hard to express, for lack of an explicit as numerous farms (see, e.g., Dalsgaard et temporal dimension). These dark clouds al. 1995), a recent, and unexpected have now been blown away, and we development. anticipate a bright future for mass-balance modelling.

91 REFERENCESa Anon. 1992. British Columbia commercial catch statistics by species, gear, month and area, 1992, Pacific Aasen, O. 1963. Length and growth of the Region. Department of Fisheries and porbeagle (Lamna nasus) in the Oceans, Vancouver. [59] north west Atlantic, Fisk. Dir. Skr. (Ser. Havunders.) 13(6): 20-37. [23] Antonellis, G.A., M.S. Lowry, e.H. Fiscus, B.S. Stewart and R.L. DeLong. 1984. Adkins, RE., R.M. Harbo and P.A. Breen. Diet of the northern elephant seal. p. 1981. A survey of commercial sea 211-233. In: B.J. LeBoeuf and R.M. urchin Strongylocentrotus francis­ Laws (eds.) Elephant Seals. canus populations in the Gulf University of California Press, Islands. Can. MS Rep. Fish. Aquat. Berkeley.[27,52] Sci. 1618, 44p. [66] Arai, M.N. 1988. Interactions of fish and Alheit, J. and T. Pitcher (eds.). 1995. Hake­ pelagic coelenterates. Can. J. ZooI. Fisheries, ecology and markets. 66: 1913-1927.[22] Chapman and Hall, London, 478p. [45] Arai, M.N. 1992. Active and passive factors affecting aggregations of Allen, KR. 1971. Relation between hydromedusae: a review. p. 99-108 production and biomass. J. Fish. Res. In: J. Bouillon, F. Boero, F. Cicogna, Board Can. 28: 1573-1581. [24, 45, J.M. Gili, & G. Hughes (eds.) 46,49] Aspects of Hydrozoan Biology. Sci. Mar. 56(2-3). Alverson, D.L, M. H. Freeberg, S.A. Murawski and J.G. Pope 1994. A Arai, M.N. and J. Fulton 1973. Diel global assessment of fisheries migration and breeding cycle of bycatch and discards. FAO Fish. Aglanthia digitale from two Tech. Pap. 339,233p. [58] locations in the Northwest pacific. J. Fish. Res. Board Can. 30:551-553. Anon. 1987. Fisheries production statistics [18] of British Columbia 1986. Ministry of Agriculture and Fisheries, Arai, M.N. and J. Fulton. 1982. Predation by Province of British Columbia, 53p. medusae on Pacific herring (Clupea [69] harengus pallasi) larvae. Can. J. Fish. Aquat. Sci. 39: 1537-1540. [65] Anon. 1988. Fisheries production statistics of British Columbia 1987. Ministry Arai, M.N. and J.e. Mason. 1982. Spring of Agriculture and Fisheries, and summer abundance and vertical Province of British Columbia, 56p. distribution of hydromedusae of the [69] central Strait of Georgia, British Columbia. Syesis 15: 7-15. [63]

a The page(s) where each reference is cited is (are) given in square brackets and thus these references may be used as an authors' index. 92 Arai, M.N. and 1.R. Jacobs. 1980. Columbia, Vancouver, B.c., 66p. [29, Interspecific predation of common 55, 70] Strait of Georgia planktonic coelenterates: laboratory evidence. Beamish, RJ. 1979. Differences in the age Can. 1. Fish. Aquat. Sci. 37: 120- of Pacific hake (Merluccius 123. [65] productus) using whole ololiths and sections of otoliths. 1. Fish. Res. Arai, M.N., G.A. McFarlane, M.W. Board. Can. 36:141-151. [46,47] Saunders and G.M. Mapstone. 1993. Spring abundance of medusae, Beamish, RJ. and D.R Bouillon. 1993. ctenophores and syphonophores off Pacific salmon production trends in southwest Vancouver Island: relation to climate, Can. 1. Fish. and possible competition or predation on Aquat. Sci. 50: 1002-1016. [12, 16] sable fish larvae. Can. Tech. Rep. Beamish, RJ. and G.A. MacFarlane. 1985. Fish. Aquat. Sci. 1939, 37p. [19] Pacific whiting, Merluccius productus, stocks off the west coats Arrhenius, F. and S. Hansson. 1993. Food of Vancouver Island, Canada. Mar. consumption of larval, young and Fish. Rev. 47: 75-81. [45,46,49] adult herring and sprat in the Baltic Sea. Mar. Ecol. Prog. Ser. 96(2): Beamish, RJ. and J.H. Youson. 1987. Life 125-137. [69] history and abundance of young adult Lampetra ayresi in the Fraser Baird, R.D., T.L. Hopkins and D.F. Wilson. River and their possible impact on 1975. Diet and feeding chronology salmon and herring stocks in the of Diaphus taaningi (Myctophidae) Strait of Georgia. Can. 1. Fish. in the Cariaco Trench. Copeia, 2: Aquat. Sci. (3): 525-537. [68] 356-65. [23] Beamish, RJ., B.L. Thomson and G.A. Barlow, J., S.L. Swartz, T.c. Eagle and P.R. MacFarlane. 1992. Spiny dogfish Wade. 1995. US Marine mammal predation on chinook and coho stock assessments: guidelines for salmon and the potential effects of preparation, background and a hatchery-produced salmon. Trans. summary of the 1995 assessments. Am. Fish. Soc. 121(4): 444-455. [69] US Dep. Commer. NOAA Tech. Memo. NMFS-OPR-6. 73p. [30] Bell, F.H. 1981. The Pacific halibut, the resource and the fishery. Alaska Barrett-Lennard, L.G., K. Heise, E. Saulitis, Northwest Publishing Co. G. Ellis and C. O. Matkin. 1995. The Anchorage, AK. [44] impact of killer whale predation on Steller sea lion populations in British Bertram, D.F. and G.W. Kaiser. 1993. Columbia and Alaska. Unpublished Rhinoceros auklet (Cerorhina report for the North Pacific monocerata) nestling diet may gauge Universities Marine Mammal Pacific sandlance (Ammodytes Research Consortium. Fisheries hexapteus) recruitment. Can. 1. Fish. Centre, University of British Aquat. Sci. 50(9): 1908-1915. [41]

93 Best, E.A. and G. St-Pierre. 1986. Pacific Brady-Campbell, M.M., D.B. Campbell and halibut as predator and prey. Tech. M.M. Harlin. 1984. Productivity of Rep. Int. Pac.Halibut Commn. kelp (Laminaria spp.) near the Seattle, IPHC 21, 27p. [41,43,44] southern limit in the northwestern Atlantic Ocean. Mar. Ecol. Prog. Ser. Bieri, R. 1961. Post-larval food of the 18: 79-88. [64] pelagic coelenterate Velella lata. Pacific Science XV: 553-556. [19] Breen, A.A. 1980. Measuring fishing intensity and annual production in Bieri, R. 1977. The ecological significance the Abalone fishery in British of seasonal occurrence and growth Columbia. Can. Tech. Rep. Fish. rate of Velella (Hydrozoa). PubI. Aquat. Sci. 947, 49p. [66] Seto Mar. BioI. Lab. 24: 63-76. [19] Brey, T. 1995. Empirische Untersuchungen Bigg, M.A. 1990. Migration of northern fur zur Populationsdynamik makro­ seal (Callorhinus ursinus), on St. Paul benthischer Evertebraten. Habilita­ Island, Alaska. Fish. Bull. 84: 383­ tionsschrift, Universitat Bremen, 394. [26, 52] 170p. [38]

Bigg, M.A., P.F. Olesiuk, G.M. Ellis, J.K.B. Brinkhurst, R.O. 1991. Benthic biology on Ford and K.C. Balcolmb III. 1990. the west Canadian continental shelf. Social organization and genealogy of Cont. Shelf Res. 11 (8-10):737-754. resident killer whales (Orcinus orca) [38] in the coastal waters of British Columbia and Washington State. Rep. Brodeur, R. 1988. Zoogeography and Int. Whal. Commn. (Special Issue) trophic ecology of the dominant 12: 383-405. [29,55] epipelagic fishes in the northern North Pacific p. 1-27 In: T. Nemoto Bonfil, R. 1994. Overview of world and W.G. Pearcy (ed.) the biology of elasmobranch fisheries. FAO the subarctic Pacific . Bull. Ocean Fisheries Tech. Paper No. 341, 119 Res. Inst. Tokyo. No. 26 (Part II). p. [24] [23,24,25]

Bonner, W.N. 1989. Whales of the world. Brodeur, R.D. and D.M. Ware. 1992. Long­ Fact on File, New York. 192 p. [26] term variability in zooplankton biomass in the subarctic Pacific Boutillier, lA. 1991. Assessment ofthe area Ocean. Fisheries Oceanography 1(1): 124 shrimp, Pandalus jordani, trawl 32-38. [16] fishery. Pacific Stock Assessment Review Committee, PSARC Adv. Brodeur, R.D. and D.M. Ware. 1995. Doc. 191-8. [39] Interdecadal variability in distribution and catch rates of Bradford, M.J. 1995. Comparative review of epipelagic nekton in the Northeast Pacific salmon survival rates. Can. l Pacific Ocean, p. 329-356. In: R.J. Fish. Aquat. Sci. 52: 1327-1338. [21] Beamish (ed.) Climate change and northern fish populations. Can. Spec. Publ. Fish. Aquat. Sci. 121. [24, 25]

94 Burgner, RL. 1991. Life history of sockeye system (version. 2.1). ICLARM salmon (Oncorhynchus nerka). p.l­ Software 6, 72 p. [12, 15,44] 118 In: C. Groot and L. Margolis (eds) Pacific salmon life histories. Christensen, V. and D. Pauly. 1995. Fish UBC Press, Vancouver. [22] production, cathces and the carrying capacity of the world ocean. Naga, Butler, RW. 1989. Breeding ecology and the ICLARM Quarterly. 18(3):34-40. population trends of the Great Blue [15, 83] Heron (Ardea Herodias Fannini) in the Strait of Georgia. p. 112-117 In: Christensen, V. and D. Pauly. 1996. K. Vermeer and R.W. Butler (eds). Ecological modelling for all. Naga, 1989. The ecology and status of the ICLARM Quarterly. 19(2). [15, marine and shoreline birds in the 90] Strait of Georgia, British Columbia. Special Publication. Canadian Christensen, V. and D. Pauly. Ecopath for Wildlife Service, Ottawa. [72, 73] Windows: a User's Guide. ICLARM, Manila. (in press). [91] Calvin, N. I. and RJ. Ellis. 1978. Quantitative and qualitative Clarke, A. and D. Morris. 1984. observations on Laminaria digitata Development ofan energy budget for and other subtidal kelps of souther E. superba, p. 102-110. In S.B. Kodiak Island, Alaska. Mar. Bioi Schnack (ed.) On the Biology of krill 47:331-336. [64] Euphausia superba. Proceedings of the seminar and Report of the Krill Campbell, R.W., N.K. Dawe, I. McTaggart­ Ecology Group, Bremerhaven, 12-16 Cowan, 1.M. Cooper, G.W. Kaiser and May 1983. Berichte zur M.C.E. McNall. 1990. The birds of Polarforschung des Alfred-Wegener­ British Columbia. Environment Instituts, Sonderheft (4). [20] Canada. [56] Compagno, LJ.V. 1984. Sharks of the Christensen, V. 1995. A model of trophic world: an annotated and illustrated interactions in the North Sea in 1981, catalogue of shark species known to the year of the stomach. Dana date. Part 1. Hexanchiformes to 11(1): 1-28. [13, 66, 76] Lamniformes. FAO Fish. Synop. (125, Vol. 4), 249 p. [24] Christensen, V. and D. Pauly (editors). 1993. Trophic models of aquatic Compton, B. D., D. Gardiner, M. Thomas, ecosystems. ICLARM Conference and 1. Michel 1994. The sucker: a Proceedings 26, 390 p. [13, 31, 66] fish full of bones, coyotes, coots and clam shell. p. 54-78 In: Collected Christensen, V. and D. Pauly. 1992a. Papers of the 29th Conference on ECOPATH II - A system for Salish and Neighbouring Languages, balancing steady-state ecosystem 11-13 August 1994, Salish Kootenai models and calculating network College, Pablo, Montana. [88] characteristics. Ecol. Modeling 61: 169-185. [12, 15] Cooney, R.T. 1987. Zooplankton. In: P.W. Hood and ST. Zimmerman (eds). Christensen, V. and D. Pauly. 1992b. A The Gulf of Alaska: Physical guide to the ECOPATH II software 95 Cooney, R.T. 1987. Zooplankton. In: P.W. Dalsgaard, J.P.T.D., C. Lightfoot and V. Hood and S.T. Zimmerman (eds). Christensen. 1995. Towards quan­ The Gulf of Alaska: Physical tification of ecological sustainability environment and biological in farming systems analysis. Ecol. resources. NOAA, Alaska. [18] Modelling 4: 181-189. [91]

Cooney, R.T. 1988. Distribution and Dark, T.A 1975. Age and growth of Pacific ecology of zooplankton in the Gulf hake, Merluccius productus. U.S. of Alaska: A synopsis. Bull. Ocean. Fish. Bull. 73: 336-355. [46] Research Inst. U. Tokyo. 26: 27-41. DeGange, AR. and G.A Sanger. 1986. [18] Marine birds. In: D.W. Hood and S.T. Zimmerman (eds.). The Gulf of Crowley, P.H. 1975. Natural selection and Alaska: Physical Environment and the Michaelis constant. 1. Theor. Biological Resources. Alaska Office, BioI. 50: 461-475. [83] Ocean Assessments Division, Canada. [31 ] Cury, P. 1994. Obstinate nature: an ecology of individuals. Thoughts on DeLong, R.L., B.S. Stewart and R.D. Hill. reproductive behavior and 1992. Documenting migrations of biodiversity. Can. J. Fish. Aquat. Sci. northern elephant seals using day 51: 1664-1673. [89] length. Mar. Marnm. Sci. 8:155-159. [27,52] Cury, P. and C. Roy. 1989. Optimal environmental window and pelagic Deriso, R.B. and TJ. Quinn II. 1983. The fish recruitment success in upwelling Pacific halibut resource and fishery areas. Can. 1. Fish. Aquat. Sci. 46: in regulatory area 2. II. Estimates of 670-680. [87] biomass, surplus production, and reproductive value. IPHC Sci. Rep. Dagg, M. 1993. Grazing by the copepod No. 67:55-89. [43] community does not control Deriso, R.B., TJ. Quinn II and P.R. Neal. phytoplankton production in the 1985. Catch-age analysis with subarctic Pacific Ocean. Prog. auxiliary information. Can. J. Oceanog. 32: 163-183. [17] Fish. Aquat. Sci. 42: 815-824. [43]

Dagg, M.L., B.W. Frost and W.E. Walser, Dill, L.M. 1983. Adaptive flexibility in the Jr. 1989. Copepod diel migration, foraging behavior of fishes. Can. J. feeding and the vertical flux of Fish. Aquat. Sci. 40: 398-408. [50] phaeopigments. Limn. Oceanogr. 34:1062-1071 [17] Dorn, M.W. 1992. Detecting environmental correlates of Pacific whiting, Dahlstrom, W.A 1970. Synopsis of Merluccius productus, growth using biological data on the ocean shrimp a growth-increment regression Pandalus jordani Rathbun, 1902. model. U.S. Fish. Bull. 90:260-275. FAO Fish. Rep. 57(4): 1377-1416. [48,49] [39]

96 Dorn, M.W., E.P. Nunnallee, C.D. Wilson, Ellis, D.V. 1968c. Quantitative benthic M. Wilkins and M.W. Saunders. investigations. V. Species data from 1993. Status of the coastal Pacific selected stations for selected stations whiting resource in 1993. p. AI-A82. (Strait of Georgia and adjacent In Pacific Fishery Management inlets), February 1965-May 1966. Council, status of the Pacific coast Fish Res. Board. Can. Tech. Rep. 73. groundfish fishery through 1993 and [65] recommended acceptable biological catches in 1994. Pacific Fishery Environment Canada. 1994. State of the Management Council, Metro Center, Environment Bulletin, September Suite 420, 2000 SW First Ave., 1994, No. 94-95, 4p. [68, 69] Portland, OR 97201. [67] Eschmeyer, W.N. E.S. Herald and H. Ellis, D.V. 1967a. Quantitative benthic Hammann. 1983. A field guide to investigations. I. Satellite Channel Pacific coast fishes of North biomass summaries and major America. Houghton Mifflin Co., taxonomy rank orders, February Boston, U.S.A., 336p. [24] 1965-May 1967. Fish Res. Board. Can. Tech. Rep. 25. [65] Evans, P.G.H. 1987. The natural history of whales and dolphins. Facts on File Ellis, D.V. 1967b. Quantitative benthic Publications, New York. 343p. [28] investigations. II. Satellite Channel species data, and February 1965-May Fargo, J. 1991. Flatfish. p. 71-115. In 1. 1967. Fish Res. Board. Can. Tech. Fargo and B.M. Leaman (eds.) Rep.38. [65] Groundfish stock assessments for the west coast of Canada in 1990 and Ellis, D.V. 1968a. Quantitative benthic recommended yield options for investigations. III. Locality and 1991. Can. Tech. Rep. Fish. Aquat. environmental data for selected Sci. 1778. 320p. [67] stations (mainly from Satellite Channel, Strait of Georgia and Fargo, J. 1995. Flatfish. p. 160-222. In M. adjacent inlets), February 1965­ Stocker and J. Fargo (eds.) December 1967. Fish Res. Boardd. Groundfish stock assessments for the Can. Tech. Rep. 59. [65] west coast of Canada in 1994 and recommended yield options for Ellis, D.V. 1968b. Quantitative benthic 1995. Can. Tech. Rep. Fish. Aquat. investigations. IV. Biomass Sci. 2069. 440p. [67] summaries and major taxon rank orders from selected stations (mainly Fasham, MJ.R. 1995. Variations in the from Satellite Channel, Strait of seasonal cycle of biological Georgia and adjacent inlets), production in subarctic oceans: A February 1965-December 1967. Fish model sensitivity analysis. Deep-Sea Res. Board. Can. Tech. Rep. 60. [65] Res. 42: 1111-1149. [86]

97 Forbes, J.R., KL. Denman, D.L. Mackas Gambell, R 1985b. Sei whale (Balaenoptera and RM. Brown. 1988. Ocean borealis Lesson 1828. p. 155-170 In: ecology data report: Subarctic S.H. Ridgway and R Harrison (eds.) Pacific Ocean , May 1984 (Project Handbook of Marine Mammals. Vol. SUPER). Can. Data Rep. 3. The sirenians and baleen whales. Hydrography Ocean Sci. 64, 147p. Academic Press. [28] [18] Gardner, G.A. 1977. Analysis of Ford, J.KB., G.M. Ellis and KC. Balcomb. zooplankton population fluctuations 1994. Killer whales: the natural in the Strait of Georgia, British history and genealogy of Orcinus Columbia. 1. Fish. Res. Bd. Can. 34: orca in British Columbia and 1196-1206. [28] Washington State. UBC Press, Vancouver. 102p. [29,55, 70] Gerrin, P.1. and M.A. Johnson. 1990. Prey identified from stomachs of harbour Froese, R. and D. Pauly. (editors). 1996. porpoise and chinook salmon from FishBase 96: Concepts, design and the 1988-89 Makah chinook salmon data sources. ICLARM, Manila, net-fishery. In: H. Kajimura (ed.) 179p. [88] Harbour porpoise abundance and interactions with the Makah chinook Frost, B.W. 1987. Grazing control of salmon net fishery in coastal phytoplankton stock in the open Washington waters, 1988-89. Unpubl. subarctic Pacific Ocean: a model Report for the National Marine assessing the role of Mammal Laboratory, Alaska mesozooplankton, particularly the Fisheries Science Center, NMFS, large calanoid copeods Neocalanus NOAA, 7600 Sand Point Way, N.E. spp. Mar. Ecol. Prog. Ser. 39: 49-68. Seattle, WA 98115. [54] [17] Gifford, D.J. 1993. Protozoa in the diets of Frost, B.W. 1993. A modelling study of Neocalanus spp. in the oceanic processes regulating plankton subarctic Pacific Ocean. Prog. standing stock and production in the Oceanog. 32: 223-237. [17] open subarctic Pacific Ocean. Prog. Oceanog. 32: 17-56. [17] Gj0saeter, 1. and K Kawaguchi. 1980. A review of the world resources of Fulton, lM.N., M. Arai and J.C. Mason. mesopelagic fish. FAO Fish. Tech. 1982. Euphausiids, coelentrates, PaperNo. 193, 151 p. [23] ctenophores, and other zooplankton from Canadian Pacific Coast Green, G.A., J.1. Brueggeman, RA. ichthyoplankton survey 1980. Can. Grotefendt, C.E. Bowlby, M.L. Rep. Fish. Aquat. Sci. No. 1125, Bonnell and K.c. Balcomb. 1992. 75p. [37] Cetacean distribution and abundance off Oregon and Washington, 1989­ Gambell, R. 1985a. Fin whale (Balaenoptera 1990. Final report for Pacific OCS physalus (Linnaeus, 1758). p. 171­ Region, Minerals Managemant 192. In: S.H. Ridgway and R. Service, California. Contract No 14­ Harrison (eds.) Handbook of Marine 12-0001-30426. lOOp. [29,53,54] Mammals. Vol. 3. The sirenians and baleen whales. Academic Press. [28]

98 Hand, e.M. and LJ. Richards. 1991. Inshore Healey, M.C. 1991. Life history of Chinook rockfish. p. 277-302. In: J. Fargo and salmon (Oncorhynchus B.M. Leaman (eds.) Groundfish tshawytscha). p. 311-394. In: e. stock assessments for the west coast Groot and L. Margolis (eds.) Pacific of Canada in 1990 and recommended salmon life histories. UBC Press, yield options for 1991. Can. Tech. Vancouver. [23] Rep. Fish. Aquat. Sci. 1778. 320p. [67] Healey, MC. 1993. The management of Pacific salmon fisheries in British Harbo, R, S. Farlinger, K. Hobbs and G. Columbia.. p. 243-266. In: L.S. Thomas. 1992. A review of quota Parsons and W.H. Lear (eds.). management in the geoduck clam Perspectives on Canadian marine fishery in British Columbia, 1976 to fisheries management. Can. Bull. 1990, and quota options for the 1991 Fish. Aquat. Sci. No 226. [21] fishery. Can. MS Rep. Fish Aquat. Sci. No. 2178. [74] Heard, W.R. 1991. Life history of pink salmon (Oncorhynchus gorbuscha). Harrison, PJ., J.D. Fulton, FJ.R. Taylor and p. 119-230. In: e. Groot and L. T.R. Parsons. 1983. Review of the Margolis (eds). Pacific salmon life biological oceanography of the Strait histories. UBC Press, Vancouver. of Georgia: pelagic environment. [22] Can. l Fish Aquat. Sci. 40: 1064­ 1094. [63, 65] Heath, W.A. 1977. The ecology and harvesting of euphausiids in the Hart, J.L. 1973. Pacific Fishes of Canada. Strait of Georgia. Ph.D. thesis, Bull. Fish. Res. Board. Can. 180: University of British Columbia, 741p. [41, 73] 187p. [65]

Hatler, D.F., RW. Campbell and A. Dorst. Heise K., G. Ellis and e.O. Matkin. 1992. A 1978. Birds of Pacific Rim National catalogue of the killer whales of Park. ace. Papers of the B.C. Provo Prince William Sound. Homer, Museum No 20, Canada. [56] Alaska: North Gulf Oceanic Society. Available from the National Marine Hay, D.E. and J. Fulton. 1983. Potential Mammal Laboratory, 7600 Sand secondary production from herring Point Way, Seattle, Washington spawning in the Strait of Georgia. 98115. 51p. [29] Can. J. Fish. Aquat. Sci. 40(2): 109­ 113. [40,68,69] Hirota, J. 1974. Quantitative natural history of Pleurobachia lochei in La Jolla Healey, M.C. 1978. The distribution, bight. U.S. Fish. Bull. 72(2): 295­ abundance, and feeding habits of 335. [40] juvenile Pacific salmon in Georgia Strait, British Columbia. Fish. Mar. Hobbs, RC. and lA. Lerzak. 1993. Servo (Can.) Tech. Rep. 788, 49p. Abundance of Pacific white-sided [23] dolphin and Dall's porpoises in Alaska estimated in the North Pacific

99 Ocean and Bering Sea during 1987 Oceanography of the Northern through 1991. Annual report to the Pacific Ocean. Idemitsu Shoten, MMPA Assessment Program, Office Tokyo. [18] of Protected Resources, NMFS, NOAA, 1335 East-West Highway, Innes, S., D.M. Lavigne, W.M. Earle and Silver Spring, MD 20910. [29] K.M. Kovacs. 1987. Feeding rates of seals and whales. 1. Anim. Ecol. Hoelzel, A.R., E.M. Dorsey and SJ. Stem. 56: 115-130. [26,29, 53, 55] 1989. The foraging specializations of individual minke whales. Anim. IPHC. 1987. The Pacific halibut: biology, Behav. 38: 786-794. [54] fishery, and management. IPHC Technical Rep. No. 22. [43] Hoenig, 1. 1983. Empirical use of longevity data to estimate mortality rates. U.S. IPHC. 1991. Annual Report, 1990. Fish. Bull. 82:898-903. [24] International Pacific Halibut Commission. Holland, A. and D. Carre. 1974. Les xanthelles des radiolaires Ivankov, V.N. 1993. Population sphaerocollides des acanthaires et de organization of Pacific salmon with Velella veliella: infrastructure­ short freshwater residence. J. cytochimie-taxonomie. Protislogica Ichthyol. 33(5): 56-65. 10(4): 573-601. [19] Jamieson, G.S. 1984. Shellfish management Hovasse, R. 1923. Endodinium chattonii advice, Pacific region. Can. MS Rep. (nov gen., nov. sp.) parasite des Fish Aquat. Sci. No. 1774. [66] Velelles un type exceptionnel de variation du nombre des Jarre-Teichmann, A. 1995. Seasonal mass­ chromosomes. Bull. BioI. Fr. Belg. balance models ofcarbon flow in the 57: 107-130. [19] central Baltic Sea with emphasis on the upper trophic levels.ICES C.M. Hughes, S.E. 1974. Stock composItIOn, 1995/T:6. [90] growth, mortality, and availability of Pacific saury, Cololabis saira, of the Jarre-Teichmann, A. and D. Pauly. 1993. northeastern Pacific Ocean. U.S. Seasonal changes in the Peruvian Fish. Bull. 72:121-131. [25] Upwelling ecosystem. p. 307-314. In: V. Christensen and D. Pauly Ignell, S.E. and J.M. Murphy. 1993. (eds). Trophic models of aquatic Salmonid spatial patterns near the ecosystems. ICLARM Conf. Proc. North Pacific Subarctic frontal zone. 26. [90] INPFC 53(2): 253-271. [21] Jarre-Teichmann, A. and V. Christensen.. Ikeda, T. 1972. Chemical composition and Comparative studies on eastern nutrition of zooplankton in the boundary current ecosystems. Bering Sea. p. 433-442. In: A.Y. ICLARM Studies and Reviews (in Takenouti (ed.) Biological press) [25]

100 Jarre-Teichmann, A., T. Brey, U.V. Ketchen, K.S. 1961. Observations on the Bathmann, C. Dahm, G.S. ecology of the Pacific cod (Gadus Dieckmann, M. Gorny, M. Klages, macrocephalus) in Canadian waters. F. Pages, J. PlOtz, S.B. Schnack­ J. Fish. Res. Bd. Canada 18: 513­ Schiel, M. Stiller and W.E. Arntz. 558. [41] 1995. Trophic flows in the benthic shelf community of the eastern Kimura, D.K. 1989. Variability in Weddell Sea, Antarctica.. In B. estimating catch in numbers-at-age Battaglia, J. Valencia and D. Walton and its impact on cohort analysis. In: (eds.) Antarctic Communities: R.J.Beamish, and G.A. MacFarlane Species, Structure and Survival. (eds) Effects of ocean variability on Cambridge University Press. recruitment and an evaluation of Cambridge, England. (in press). [66] parameters used in stock assessment models. Can. Spec. Publ. Fish. Jefferson, T.A. 1990. Status of Dall's Aquat. Sci. 108: 57-66. [47] porpoise, Phocoenoides dalli, in Canada. Canadian Field Naturalist Kirchman, D.L., R.G. Kiel, M. Simon, and 104: 112-116. [28] N.A. Welschmeyer. 1993. Biomass and production of heterotrophic Jefferson, T.A., S. Leatherwood and M.A. bacterioplankton in the oceanic Webber. 1993. Marine mammals of subarctic Pacific. Deep-Sea Res. 40: the world. Rome, FAO. 320p. [27,30] 967-988. [16]

Johannes, R.E. 1981. Words of the lagoon: Klinowska, M. 1991. Dolphins, porpoises and fishing and marine lore in the Palau whales of the world. The IUCN red district of Micronesia. University of data book. IUCN, Gland, Switzerland California Press, Berkeley, 245p. and Cambridge U.K. 429p. [29, 30] [88] Kozloff, E.N. 1987. Marine invertebrates of Jones, B.C. and G.H. Geen. 1977. Food and the Pacific Northwest. University of feeding of spiny dogfish (Squalus Washington Press, Seattle. 511 p. acanthias) in British Columbia [38] waters. J. Fish. Res. Board. Can. 34: 2067-2078. [45,69] Laevastu, T. and H.A. Larkins 1981. Marine fisheries ecosystems: its quantitative Kasahara, H. 1961. Fisheries resources of evaluation and management. Fishing the North Pacific Ocean. Part 1. H.R. News Books, Farnham, Surrey, MacMillan Lectures in Fishreries. England, 160p. [12] Univ. British Columbia, Vancouver, Canada. 135p. [25] Landry, M.R., D.J. Gifford, D.L. Kirchman, P.A. Wheeler and B.C. Monger. Keiser, R. 1982. Hydroacoustic biomass 1993. Direct and indirect effects of estimates ofbathypelagic ground fish grazing by Neocalanus plumchrus on in Georgia Strait, January, February plankton community dynamics in the and April, 1981. Can. MS Rept. Fish subarctic Pacific. Prog. Oceanog. 32: Aquat. Sci. No. 1715. [68] 239-258. [17]

101 Larkin, P.A. 1992. Analysis of possible LeBoeuf and R.M. Laws (eds.) causes of the shortfall in sockeye Elephant Seals. University of spawners in the Fraser River. California Press, Berkeley. [27,52] Technical Appendix to "Managing salmon on the Fraser" by Peter H. LeBrasseur, R.J. 1966. Stomach contents of Pearse. 33p. salmon and steelhead trout in the Northeastern Pacific Ocean. J. Fish. Larson, RJ. 1986. Seasonal changes in the Res. Board Can. 32(1): 85-100. [23] standing stocks, growth rates, and production rates of gelatinous Legare, lE.H. 1957. The qualitative and predators in Saanich Inlet, British quantitative distribution of plankton Columbia. Mar. Ecoi. Prog. Ser. 33: in the Strait of Georgia in relation to 89-98. [39,40, 65] certain oceanographic factors. J. Fish. Res. Bd. Can. 14(4): 521-552. Larson, R.J. 1987a. Trophic ecology of [63] planktonic gelatinous predators in Saanich Inlet, British Columbia: Levings, C.D., R.E. Foreman, and V.J. diets and prey selection. J. Plankton Tunnicliffe. 1983. Review of the Res. 9(5): 811-820. [65] benthos of the Strait of Georgia and contiguous fjords. Can. J. Fish. Larson, RJ. 1987b. Daily ration and Aquat. Sci. 40: 1120-1141. [65,66] predation by medusae and ctenophores in Saanich Inlet, British Lindley, lA. 1980. Population dynamics Columbia, Canada. Netherl. J.Sea and production of euphausiids. II. Res. 21(1): 35-44. [65] Thysanoessa inermis and T raschi in the North Sea and American coastal Lasker, R. 1966. Feeding, growth, waters. Mar. Bioi. 59: 225-233. [20] respiration and carbon utilization ofa euphausiid crustacean. l Fish. Res. Livingston, P.A. & Bailey, K.M. 1985. Bd. Can. 23: 1291-1317. [20] Trophic role of the Pacific whiting, Merluccius productus. Mar. Fish. Lawson, RJ. 1986. Water content, organic Rev. 47: 16-22. [49] content and carbon and nitrogen composition of medusae from the Livingston, P.A. 1983. Food habits of northeast Pacific. J. expo Mar. Bioi. Pacific whiting, Merluccius Ecoi. 99: 107-120. productus, off the west coast of North America, 1967 and 1980. U.S. Leatherwood, S., R. Reeves, W.F. Perrin and Fish. Bull. 81 :629-636. [50] W.E. Evans. 1982. Whales, dolphins and porpoises of the eastern north Lockyer, C. 1981 a. Estimates of growth and Pacific and adjacent Arctic waters. energy budget for the sperm whale, NOAA Tech. Rep. NMFS Circular Physeter catodon. p. 379-488. In: 444. 245p. [27] FAO Advisory Committee in Marine Resources Research Working Party LeBoeuf, BJ. 1994. Variation in the diving on Marine Mammals Mammals in the pattern ofnorthern elephant seals with seas. Volume 3. General papers and age, mass, sex and reproductive large cetaceans. [28] condition. p. 237-252. In: BJ.

102 Lockyer, C. 1981b. Growth and energy Mauchline, J. 1980. The biology of budgets of large baleen whales from euphausiids. Adv. Mar. BioI. 18: the Southern Hemisphere. p. 379-488. 373-623. [20] In: FAO Advisory Committee in Marine Resources Research Working Mauchline, J. and L.R. Fisher. 1969. The Party on Marine Mammals Mammals biology of euphausiids. Adv. Mar. in the seas. Volume 3. General papers BioI. 7, 454 p. [20] and large cetaceans. [27,28,53] McAllister, M., E.K. Pikitch, A.E. Punt and Luning, K. 1969. Standing crop and leaf R. Hilborn. 1994. A Bayesian area index of the sublittoral approach to stock assessment and Laminaria species near Helgoland. harvest decisions using the Mar. BioI. 3:282-286. [64] sampling/importance resamp1ing algorithm. Can. J. Fish. Aquat. Sci. Mackas, D.L. 1992. Seasonal cycle of 51: 2673-87. [85,90] zooplankton off Southwestern British Columbia. 1979-1989. Can. McCall. R.A. and R.M. May 1995. More J. Fish. Aquat. Sci. 49: 903-921. [37, than a seafood platter: review of 39] FishBase - a biological database on fish. Nature 376:735. [88] Mackay, A. 1981 The generalized inverse. Practical computing, September McFarlane, G.A. and RJ. Beamish. 1983a. 1981:108-110. [15] Preliminary observations on the juvenile biology of sablefish Madin, L.P. and J.E. Purcell. 1992. Feeding, (Anoplopoma fimbria) in waters off metabolism, and growth of the west coast ofCanada, p. 119-136. Cyclosalpa bakeri in the subarctic In: Proceedings of the Int. Sablefish Pacific. LimnoI. Oceanogr. 37: 1236­ Symposium. Alaska Sea Grant Rep. 1251. [18] 83-8. [42]

Mann, K.H. 1972. Ecological energetics of McFarlane, G.A. and RJ. Beamish. 1983b. the seaweed zone in a marine bay on Biology of adult sablefish the Atlantic coast of Canada. I (Anoplopoma fimbria) in waters off Zonation and biomass of seaweeds. western Canada, p. 59-80. In: Mar. BioI. 12:1-12. [64] Proceedings of the Int. Sablefish Symposium. Alaska Sea Grant Rep. Marlow, C.G. and C.B. Miller. 1974. The 83-8. [42] vertical distribution of zooplankton at Ocean Station "P" in June-July, McFarlane, G.A. and RJ. Beamish. 1992. 1971. Oregon State University, Climatic influences linking copepod School of Oceanography Data report production with strong year classes 61. Ref. 74-13. [18] in sablefish Anoploma fimbria. J. Fish. Aquat. Sci. 49: 743-753. [16, 42]

103 Merrick, RM. 1994. Status review of Steller Limnol. Oceanogr. 36: 1600-1615. sea lions (Eumetopias jubatus). [17] Unpubl. Report for the National Marine Mammal Laboratory, Alaska Mills, C.E. 1981. Seasonal occurrence of Fisheries Science Center, NMFS, planktonic medusae and ctenophores NOAA, 7600 Sand Point Way, N.E. in the San Juan Archipelago (N.E. Seattle, WA 98115. 61p. [27] Pacific). The Wasmann Journal of' Biology 39(1-2): 6-29. [40] . Merrick, RM. 1995. The relationship of the foraging ecology of Steller sea lions Mottet, M.G. 1976. The fishery biology of (Eumetopias jubatus) to their sea urchins in the family population decline, in Alaska. Ph.D. Strongylocentrotidae. Wash. Dept. Thesis. University of Washington. Fish. Tech. Rep. No. 20, 66p. [66] 171p. [27] Mottet, M.G. 1978. A review of the fishery Methot, R.D. 1990. Synthesis model: an biology of Abalones. Wash. Dept. adaptable framework for analysis of Fish. Tech. Rep. No. 37. [66] diverse stock assessment data. Int. North Pac. Fi3h. Comm. Bull. 50: Muck, P. and D. Pauly. 1987. Monthly 259-277. [47] anchoveta consumption of Guano Birds, 1953 to 1982. p. 219-233. In: Methot, R.D. and M.W. Dom. 1995. D. Pauly and I. Tsukayama (eds.). Biology and fisheries of north The Peruvian Anchoveta and its Pacific hake (M productus). p. 389­ upwelling ecosystem: three decades 414. In: J. Albeit and T.J.Pitcher of change. ICLARM Studies and (eds) Hake: fisheries, ecology and Reviews 15. [56, 71] markets. Chapman & Hall, London. 478p. [46,47,48] Mullin, M. 1969. Production of zooplankton in the ocean: the present status and problems. Ocean. Mar. BioI. Ann. Miller, C.B. 1993. Pelagic production in the Rev. 7: 293-314. [20] Subarctic Pacific. Progr. Oceanogr. 32: 1-15. [36] Nakano, H. and Y. Watanabe 1992 Effect of high seas driftnet fisheries on blue shark stock in the North Pacific. ~ller, C.B. and SUPER Group. 1988. Compendium of documents Lower trophic level production submitted to the Scientific Review of dynamics in the oceanic Subarctic North Pacific Highseas Driftnet Pacific Ocean. Bull. Ocean. Res. Fisheries, Sidney, B.C., Canada, Vol. Inst. Univ. Tokyo 26: 1-26. [16] 1, 15p. [23]

Miller, C.B., B.W. Frost, P.A. Wheeler, Nemoto, T. 1959. Food of baleen whales in M.R. Landry, N. Welschmeyer and the northern Pacific. Sci. Rep. T.M. Powell. 1991. Ecological Whales. Res. Inst. 14: 149-290. [27, dynamics in the subarctic Pacific, a 28] possibly iron-limited ecosystem.

104 Nemoto, T. and A. Kawamura. 1977. Nuechterlein, G.L. and R.W. Storer. 1989. Characteristics of food habits and Mate feeding by Western and Clark's distribution of baleen whales and Grebes. The Condor. 91 :37-42. [72] special reference to the abundance of north Pacific Sei and Bryde's whales. Nyblade, C.F. 1979. Five-year intertidal Rep. Int. Whal. Commn. (Special community change, San Juan Issue 1): 80-87. [28] Islands, 1974-1978 and The intertidal benthos of north Puget Nichol, L. and K. Heise. 1992. The historical Sound, Summer 1978. Appendix F. occurrence of large whales off the Baseline Study Program. Queen Charlotte Islands. Unpubl. Washington State Dept. of Ecology, report for the South Moresby! Gwaii Olympia, Washington. [65] Haanas National Park Reserve. Canadian Parks Service, P.O. Box 37, Olesiuk, P.F., M.A. Bigg , G.M. Ellis, SJ. Queen Charlotte City, B.C. VOS ISO, Crockford and RJ. Wigen. 1990c. An 68 p. [27,28,54] assessment of feeding habits of harbour seals (Phoca vitulina) in the Nilsson, S.G. and LN. Nilsson. 1976. Strait of Georgia, British Columbia, Number, food and consumption, and based on scat analysis. Can. Tech. fish predation by birds in Lake Rep. Fish. Aquat. Sci. No. 1730. Mockeln, Southern Sweden. Ornis. 135p. [53, 70] Scand. 7:61-70. [56, 71] Olesiuk, P.F., M.A. Bigg and G.M. Ellis. Noakes, 1992. On growth and mortality of 1990a. Life history and population geoduck clams (Panope abrupta) In: dynamics of resident killer whales Thomas 1992. Can. MS Rep. Fish. (Orcinus orca) in the coastal waters Aquat. Sci. No. 2169. [66] of British Columbia and Washington State. Rep. Int. Whal. Commn. Northcote, T.G. and P.A. Larkin. 1989. the (Special Issue) 12: 209-243. [29, 55] Fraser River: A major salmonine production system, p. 172-204. In: Olesiuk, P.F., M.A. Bigg and G.M. Ellis. D.P. Dodge (ed.) Proceedings of the 1990b. Recent trends in the International Large River abundance of harbour seals, Phoca Symposium (LARS). Can. Spec. vitulina, in British Columbia. Can. 1. Publ. Fish. Aquat. Sci. 106. [68,69] Fish. Aquat. Sci. 47: 992-1003. [52]

Northridge, S.P. 1991. An updated world Olivieri, R.A., A. Cohen, F.P. Chavez. 1993. review of interactions between marine An ecosystem model of Monterey mammals and fisheries. FAO Bay, California. p. 315-322. In: V. Fisheries Technical Paper. No. 251 Christensen and D. Pauly (eds). Suppl. 1. Rome, FAD. 58 p. [25, 30, Trophic models of aquatic 51] ecosystems. ICLARM Conf. Proc. 26. [64]

105 Osborne, RJ., 1. Calambokidis and E.M. Parsons, T. R. 1980 Is the Strait of Georgia Dorsey. 1988. A guide to marine becoming eutrophic? Can. J. Fish. mammals of greater Puget Sound. Aquat. Sci. 37:1043-1047. Island Publishers, Anacortes, WA. 191 p. [70] Parsons, T.R., RJ. LeBrasseur and W.E. Banaclough. 1970. Levels of Outram, D.N. and C. Haegele. 1972. Food of production in the pelagic Pacific hake (Merluccius productus) environment of the Strait of Georgia, on an offshore bank southwest of B.C.: a review. J. Fish. Res. Board Vancouver Island, British Columbia. Can. 27: 1257-1264. [64] 1. Fish. Res. Bd. Can. 29: 1792­ 1795. [49] Paul, A.J., J.M. Paul and R.L. Smith. 1990. Consumption, growth, and Palmer, R.S. (editor). 1962. Handbook of evacuation in the Pacific cod, Gadus North American Birds. Volume 1. macrocephalus. J. Fish. BioI. Yale University Press, New Haven & 37:117-124. [41] London. [31,56,71,72] Pauly, D. 1980. On the interrelationships Palmer, R.S. (editor). 1976a. Handbook of between natural mortality, growth North American Birds. Volume 2. parameters, and mean environmental Yale University Press, New Haven & temperature in 175 fish stocks. London. [56, 71, 72] Journal du Conseil international pour l'Exploration de la Mer 39(3):175­ Palmer, R.S. (editor). 1976b. Handbook of 192. [45, 90] North American Birds. Volume 3. Yale University Press, New Haven & Pauly, D. 1989. Food consumption by London. [56, 71, 72] tropical and temperate fish populations: some generalizations. 1. Palmer, R.S. (editor). 1988. Handbook of Fish BioI. 35 (Suppi. A); 11-20. [67, North American Birds. Volume 4. 90] Yale University Press, New Haven & London. [56] Pauly, D. and V. Christensen. 1993. Stratified models of large marine Palomares, M.L., R. Froese and D. Pauly ecosystems: a general approach and 1993. On traditional knowledge, fish an application to the South China and databases: a call for Sea, p. 148-174. In: K. Sherman, contributions. South Pacific L.M. Alexander and B.D. Gold (eds.) Commission/Traditional Manage- Stress, mitigation and sustainability ment and Knowledge Information of large marine ecosystems. AAAS Bulletin (Noumea, New Caledonia) Press, Washington, D.C. [13] 2:17-19. [89] Pauly, D. M.L. Palomares, and R. Froese Palomares, M.L.D. and D. Pauly. 1989. A 1993a. Some prose on a database of multiple regression model for indigenous knowledge on fish. predicting the food consumption of Indigenous Knowledge and marine fish populations. Aust. 1. Development Monitor 1(1):26-27. Mar. Freshwat. Res. 40:259-273. [89] [25]

106 Pauly, D., A Trites, E. Capuli and V. Mammal Laboratory, Sand Point Christensen. 1995. Diet composition Way, Seattle WA). [26,52] and trophic levels of marine mammals.. ICES CM 1995, Marine Perez, M.A and M.A. Bigg. 1986. Diet of mammal CtteelN:13, 22 p. [27, 70] northern fur seals (Callorhinus ursinus) off western North America. Pauly, D., M. Soriano-Bartz and M.L. Fishery Bulletin 84:957-971. [26,52] Palomares. 1993b. Improved construction, parametrization and Pitcher, T.J. and 1. A1heit. 1995. What interpretation of steady-state makes a hake? A review of the ecosystem models. p.l-13. In: V. critical biological features that Christensen and D. Pauly (eds.) sustain global hake fisheries. p. 1-14. Trophic models of aquatic In: J. Alheit and T.1.Pitcher (eds) ecosystems. ICLARM Conf. Proc. Hake: fisheries, ecology and 26. [24, 60, 77] markets. Chapman & Hall, London. [46] Payne, ALL. and AE. Punt. 1995. Biology and fisheries of South African Cape Polovina, J.J. 1984. Models of a coral reef hakes (M capensis and M ecosystem I: the ECOPATH model paradoxus). p. 15-48. In: 1. Alheit and its application to French Frigate and T.1. Pitcher (eds) Hake: Schoals. Coral Reefs 3(1): 1-11. [12, fisheries, ecology and markets. 15] Chapman & Hall, London. 478p. [48] Polovina, J.1. 1985 An approach to estimating an ecosystem box model. Pearcy, W.G. 1992. Ocean ecology of north U.S. Fish. Bull. 83(3):457-560. [15] Pacific salmonids. Washington Sea Grant Program. Seattle: 179 p. [21] Polovina, 1.J. and M.D. Ow 1983. ECOPATH: a user's manual and Pearcy, W.G., 1.P. Fisher and M.M. program listings. NMFSINOAA, Yoklavich. 1993. Biology of the Honolulu Admin. Rep. H-83-23, Pacific pomfret (Brama japonica) in 46p. [15] the North Pacific Ocean. Can. 1. Fish. Aquat. Sci. 50:2608-2625. [24] Polovina, J.1., G.T. Mitchum, and GT. Evans. 1995. Decadal and basin­ Perez, M.A 1990. Review of marine scale variation in mixed layer depth mammal population and prey and the impact on biological information for Bering Sea ecosystem production in the Central and North studies. NOAA Tech. Memo. NMFS Pacific, 1960-88. Deep-Sea Res. 42 FINWC-186. US Dep. Comm. [26] (10): 1701-1716. [13, 16]

Perez, M.A. and M.A. Bigg. 1981. An Polovina, J.1., G.T. Mitchum, N.E. Graham, assessment of the feeding habits of M.P. Craig, E.E. DiMaritni, and E.N. the northern fur seal in the eastern Flint. 1994. Physical and biological North Pacific Ocean and eastern consequences of a climate event in Bering Sea. 140 p. (unpubl. report, the North Pacific. Fisheries available from National Marine Oceanography 3(1): 15-21. [12,13]

107 Pope, J.G. and J.G. Shepherd. 1982. A Pacific. Mar. Ecoi. Prog. Ser. 73: Simple Method for the Consistent 211-217. [18] Interpretation of Catch-at-Age Data. 1 Cons. Int Explor. Mer.-40: 176­ Quinn, T.1. II, R.B. Deriso and S.H. Hoag. 184. [47] 1985. Methods of population assessment of Pacific halibut. IPHC Purcell, lE. 1981. Dietary composition and Sci. Rep. No. 72. [43] diel feeding patterns of epipeiagic siphonophores. Mar. BioI. 65: 83-90. Reilly, R.1. and 1 Barlow. 1986. Rates of increase in dolphin population size. Purcell, lE. 1989. Predation of fish larvae U.S. Fish. Bull. 84: 527-533. [27, 28, and eggs by the hydromedusa 29,30,53,54,55] Aequorea victoria at a herring spawning ground in British Rexstad, E.A. and E.K. Pikitch. 1986. Columbia. Can. J. Fish. Aquat. Sci. Stomach content and food 46: 1415-1427. [65] consumption estimates of Pacific hake, Merluccius productus. Fish. Purcell, lE. 1990. Soft-bodied zooplankton Bull. 84: 947-956. [ 50] predators and competitors of larval herring (Clupea harengus pallasi) at Rice, D.W. 1986. Beaked Whales. p. 102­ herring spawning grounds in British 109. In: D. Haley (ed.) Marine Columbia. Can. J. Fish. Aquat. Sci. Mammals of eastern North Pacific 47: 505-515. [40] and Arctic waters. Second Edition. Pacific Search Press: Seattle. [30] Purcell, lE. 1991 a. Predation by Aequorea victoria on other species of Richards, L.J. and K.L. Yamanaka. 1992. potentially competing pelagic Lingcod. p. 23-55. In: B.M. Leaman hydrozoans. Mar. Ecoi. Prog. Series (ed.) Groundfish stock assessments 72: 255-260. [65] for the west coast of Canada in 1991 and recommended yield options for Purcell, lE. 1991 b. A review of cnidarians 1992. Can. Tech. Rep. Fish. Aquat. and ctenophores feeding on Sci. 1866. 304p. [67] competitors in the plankton. Hydrobiologia 216/217: 335-342. Ricker, W.E. 1976. Review of the rate of [65] growth and mortality of Pacific salmon in salt water, and noncatch Purcell, lE. and J.H. Grover. 1990. mortality caused by fishing. J. Fish. Predation and food limitation as Res. Board Can. 33(7): 1483-1524. causes of mortality in larval herring [21] at a spawning ground in British Columbia. Mar. Ecol. Prog. Ser. 59: Ricker, W.E. 1989. History and present state 55-61. [18] of the odd-year pink salmon runs of the Fraser River region. Can. Tech. Purcell, J.E. and L.P. Madin. 1991. Diel Rep. Fish. Aquat. Sci. 1702, 43p. patterns of migration, feeding, and [68] spawning by salps in the subarctic

108 Robertson, I. 1973. Predation by fish-eating Sarmiento, lL., R.D. Slater, MJ.R. Fasham, birds on stocks ofthe Pacific herring, H.W. Ducklow, J.R. Toggweiler, and Clupea pallasi, in the Gulf Islands of G.T. Evans. 1993. A seasonal three­ British Columbia. Unpubl. Rep., dimensional ecosystem model of Pacific Biological Station, Nanaimo, nitrogen cycling in the North B.e., 29p. [72] Atlantic euphotic zone. Global Biogeochemical Cycles 7: 417-450. Robinson, C.L. D.M. Ware and T.R. [86] Parsons. 1993. Simulated annual plankton production m the Saunders, M.W. 1988. Dogfish, p. 151-158. northeastern Pacific coastal In: 1 Fargo, M.W. Saunders, and upwelling domain. J. Plankton Res. A.V. Tyler (eds) Groundfish stock 15(2):161-183. [37] assessments for the West Coast of Canada in 1987 and recommended Romaine, S.l, D.L. Mackas, M.e. Macaulay yield options for 1988. Can. Tech. and DJ. Saxby. 1995 Comparisons Rep. Fish. Aquat. Sci. 1617. [45] of repeat acoustic surveys in Jervis Inlet, British Columbia: 1994-1995. Saunders, M.W. and G.A. McFarlane. 1994. Paper presented at the workshop on Pacific hake. p. 154-203. In: M. "Harvesting krill: ecological impact, Stocker (ed.) Groundfish stock assessment, products and markets", assessments for the west coast of November 14-16, Fisheries Centre, Canada in 1993 and recommended UBC, Vancouver, Canada. [40] yield options for 1994. Can. Tech. Rep. Fish. Aquat. Sci. 1975. [67] Salo, E.O. 1991. Life history of chum salmon (Oncorhynchus keta). p. 231­ Saunders, M.W., G.A. McFarlane, and W. 310 In: Groot C. and L. Margolis Shaw. 1989. Delineation of walleye (eds) Pacific salmon life histories. pollock (Theragra chalcogramma) UBC Press, Vancouver. [22] stocks off the Pacific coast of Canada. p. 379-401. In: Proceedings Sambilay, V. 1993. Proximate analysis of of the international symposium on selected organisms important for the biology and management of constructing trophic models of walleye pollock, Anchorage, Alaska, ecosystems. p. 353-355 In: V. USA, November 14-16, 1988. Christensen and D. Pauly (eds.) Alaska Sea Grant Report No. 89-1. Trophic models of aquatic [67] ecosystems. ICLARM conference Proceedings 26. [37,38] Savilov, A.I. 1961. The distribution of different forms of Velella lata Ch. Sanger, G. A. 1987. Trophic levels and and Eys. and Physalia utriculus (La trophic relationships of seabirds in Martiniere) Esch. in the North the Gulf of Alaska. p. 229-257. In: Pacific. Trud. Inst. Okeanolog. lP. Croxall (ed.) Seabirds: Feeding 45:223-239 (in Russian, with English Ecology and Role in Marine abstract). [19] Ecosystems. Cambridge University Press, New York. [31]

109 Savinykh, V.F. and L-P. Vlasova. 1994. management of walleye pollock, Size-age structure and growth rate of Anchorage, Alaska, USA, November Pacific Pomfret. Vopr. Ikhtiol. 34: 14-16, 1988. Alaska Sea Grant 494-500. [24] Report No. 89-1. [67]

Scammon, C.M. 1874. The marine mammals Shenker, 1.M. 1984. Scyphomedusae in of the north-western coast of North surface waters near the Oregon America. Reprinted by Dover: New Coast, May-August, 1981. Estuarine York. 319p. [27] Coastal and Shelf Sci. 19: 619-632.

Schnack, S.B., V. Smetacek, V., B. von Shimazaki, K. 1989. Ecological studies of Bodungen, and P. Stegmann 1985. the pomfret (Brama japonica) in the Utilization of phytoplankton by North Pacific Ocean, p. 195-205. In: copepods in Antarctic waters during RJ. Beamish and G.A. McFarlane spring, p. 65-81. In: J.S. Gray and (ed.) Effects of ocean variability on M.E. Christiansen (eds.) Marine recruitment and an evaluation of biology of polar regions and effects parameters used in stock assessment of stress on marine organisms. John models. Can. Spec. Publ. Fish. Wiley & Sons. [20] Aquat. Sci. 108. [24]

Schweigert, 1.F. and C. Fort. 1994. Stock Siegel, V. 1986. Untersuchungen zur assessment for British Columbia Biologie des antarktischen Krills, herring in 1993 and forecast of the Euphausia superba, im Bereich der potential catch in 1994. Can. Tech. Bransfield StraBe und angrenzender Rep. Fish. Aquat. Sci. 1971. 66p. Gebiete. Mitt. Inst. Seefisch., [68] Hamburg (38), 244p. [20]

Schweigert, 1.G. 1988. Status of the Pacific Simon, M., N.A. Welschmeyer, and D.L. sardine, Sardinops sagax, in Canada. Kirchman. 1992. Bacterial Can. Field. Nat. 102(2): 296-303. production and the sinking flux of [41 ] particulate organic matter in the subarctic Pacific. Deep-Sea Res. 39: Seki, H., K.V. Stevens and T.R. Parsons 1997-2008. [16] 1969 the contribution of allochthonous bacteria and organic Small, RJ. and D. DeMaster. 1995. Alaska material from a small river into a marine mammal stock assessments semi enclosed sea. Arch. Hydorbiol. 1995. US Dept. Commer., NOAA 66(1):37-47. [64] Tech. Memo. NMFS-AFSC-57. 93p. [26,27,28,29,51,53,55] Shaw, W., G.A. McFarlane, and M.W. Saunders. 1989. Distribution and Smith, B.D. and G.S. Jamieson. 1991. abundance of walleye pollock Movement, spatial distribution and (Theragra chalcogramma) in the mortality of male and female Strait of Georgia based on estimates Dungeness crabs Cancer magister in 1981 and 1988. p. 349-377. In: near Tofino, British Columbia. Fish. Proceedings of the international Bull. 89: 137-148. symposium on the biology and

110 Smith, B.D., G.A. McFarlane and AJ. Casso Stewart, B.S. and RL. DeLong. 1995. 1990. Movements and mortality of Double migrations of the northern tagged male and female lingcod in elephant seal Mirounga the Strait of Georgia, British angustrirostris. 1. Mamm. 76: 196­ Columbia. Trans. Am. Fish. Soc. 205. [27,52] 119: 813-824. [67] Stewart, B.S. and S. Leatherwood. 1985. Smith, B.D., M.W. Saunders and G.A. Minke Whale Balaenoptera MacFarlane. 1992. Use of binomial­ acutorostrata Lacepede 1804. p. 91­ normal functional model to 136. In: S.H. Ridgway and R. demonstrate that the sex and age Harrison (eds.) Handbook of Marine composition of Pacific hake Mammals. Vol. 3. The sirenians and (Merluccius productus) aggregations baleen whales. Academic Press. [54] affects estimates of mean lengths-at­ age. Can. 1. Fish. Aquat. Sci. 49: Stewart, B.S., R.K. Yochem, H.R Huber, 1657-1669. [45] RL. DeLong, R1. Jameson, WJ. Sydeman, S.G. Allen, and BJ. Soutar, A. and Isaacs 1. D. (1974) LeBoeuf. 1994. History and present Abundance of pelagic fish during the status of the northern elephant seal 19th and 20th centuries as recorded population. p. 29-48. In: BJ. LeBoeuf in anaerobic sediment off the and RM. Laws (eds.) Elephant Seals. Californias. Fish. Bull. 72: 257-273. University of California Press, [46] Berkeley. [27,52]

St-Pierre, G. 1989. Recent studies of Pacific Stillwell, C.E. and N.E. Kohler. 1982. Food, halibut postlarvae in the Gulf of feeding habits and estimates of daily Alaska and eastern Bering Sea. ration of the shortfin mako (lsurus IPHC Sci. Rep. No. 73. [43] oxyrinchus) in the northwest Atlantic. Can. J. Fish. Aquat. Sci. 39: Stauffer, G.D. 1985. Biology and history of 407-414. [24] the coastal stock of Pacific whiting. Merluccius productus. Mar. Fish. Stocker, M. (editor). 1994. Groundfish stock Rev. 47: 2-7. [45] assessments for the West coast of Canada in 1993 and recommended Stevens, G., D.A. Armstrong and R. yield options for 1994. Can. Tech. Cusimano. 1982. Feeding habits of Rep. Fish. Aquat. Sci. 1975, 352p. the Dungeness Crab Cancer magister [41,42,45] as determined by the index of relative importance. Mar. BioI. 72: Stocker, M. 1993. Recent management of 135-145. [66] the British Columbia herring fishery, p. 267-293. In: L.S. Parsons and Stewart, B.S. and R.L. DeLong. 1994. W.H. Lear (eds.) Perspectives on Postbreeding foraging migrations of Canadian marine fisheries northern elephant seals p. 290-309. management. Can. Bull. Fish. Aquat. In: BJ. LeBoeuf and R.M. Laws Sci. 226. [40,68,69] (eds.) Elephant Seals. University of California Press, Berkeley. [52]

III Stocker, M., C. Hand, and M.K. McAllister. "Harvesting krill: ecological impact, 1995. Pacific cod. p. 115-159. In M. assessment, products and markets", Stocker and J. Fargo (eds.) November 14-16, Vancouver, Groundfish stock assessments for the Canada. [20, 40] west coast ofCanada in 1994. [67] Tanasichuk, R.W., D.M. Ware, W. Shaw Stockner, 1.G., D.D. Cliff and K.R.S. and MacFarlane, G.A. 1991. Shortreed. 1979. Phytoplankton Variations in diet, daily ration, and ecology the Strait of Georgia, British feeding periodicity of Pacific hake Columbia. J. Fish. Res. Board Can. (Merluccius productus) and spiny 36: 657-666. [63,64] dogfish (Squalus acanthias) off the lower west coast of Vancouver Strom, S.L., J.R. Postel, and B.C. Booth. Island. Can. J. Fish. Aquat. Sci. 48: 1993. Abundance, variability, and 2118-2128. [49,50,51,69] potential grazing impact of planktonic ciliates in the open Taylor, D.L. 1971. Ultrastructure of the subarctic Pacific Ocean. Prog. "zooxanthella" Endodinium chattonii Oceanog. 32: 185-203. [17] in situ. 1. mar. BioI. Assoc U.K. 51: 227-234. [19] Stroud, R.K., C.H. Fiscus and H. Kajimura. 1981. Food of the Pacific white-sided Terazaki, M. and C.B. Miller. 1986. Life dolphin, Lagenorhynchus history and vertical distribution of obliquidens, Dall's porpoise, pelagic chaetognaths at Ocean Phocoenoides dalli, and northern fur Station P in the subarctic Pacific. seal Callorhinus ursinus, off Deep-Sea Research 33(3): 323-337. California and Washington. U.S. Fish. [18] Bull. 78(4): 951-959. [54] Thomson, B.L. 1994. Spiny dogfish, p. 204­ Sullivan, B.K. 1980. In situ feeding 217. In: M. Stocker (ed.) Groundfish behaviour at Sagitta elegans and stock assessment for the west coast Eukrohnia hamata (Chaetognatha) in ofCanada in 1993 and recommended relation to the vertical distribution yield options for 1994. Can. Tech. and abundance of prey at ocean Rep. Fish. Aquat. Sci. 1975. [68,69] Station "P". Limnol. Oceanogr. 25: 317-326. [18] Tokranov, A.M., A.B. Vinnikov. 1991. Diet of the Pacific cod (Gadus morhua Swartzman, G.L., W.M. Getz and R.C. macrocephalus) and its position in Francis. 1987. Binational the food chain in Kamchatkan management of Pacific hake coastal waters. Journal of (Merluccius productus): a stochastic Ichthyology 311, No.5. [41] modeling approach. Can J.Fish Aquat. Sci. 44:1053-1063. [48] Tomilin, A.G. 1957. Cetacea. Vol. 9 In: V.G. Heptner (ed.) Mammals of the USSR Tanasichuk, R. 1995. The importance of and adjacent countries. Izd. Akad. euphausiids to a British Columbia Nauk SSSR, Moscow. 756 p. (Trans!. coastal marine ecosystem. Paper from Russian by Israel Prog. Sci. presented at the workshop on Trans!., 1967, 717p). [30,53]

112 Townsend, C.H. 1935. The distribution of Turnock, B.J. and T.J. Quinn. 1991. The certain whales as shown by logbook effect of responsive movement on records of American whaleships. abundance estimation using line Zoologica. (N.y.) 19: 1-50. [27,28] transect sampling. Biometrics 47: 701-715. [29] Treacy, S.D. 1985. Feeding habits of marine mammals from Grays Harbour Tutubalin, G.B. and V.I. Chuchukalo. 1992. Washington to Netarts Bay, Oregon. The feeding of Oncorhynchus Pacific In Marine mammals and their salmons in the North Pacific during interactions with fisheries of the the winter-spring period p. 77-85. In: h Gritsenko (ed.) Living resources of Columbia River and adjacent waters the Pacific Ocean. Biologicheski 1980-1982. Third Annual Report to Resursy Tikhogo Okeana., Moskva the National Marine Fisheries (in Russian, with English abstract). Service, Seattle, WA [54] [68,69]

Trites, AW. and D. Pauly Estimating the Tyler, A.V. and R.P. Foucher. 1990. Pacific mean (population-weighted) biomass cod. In: A.V. Tyler and J. Fargo of marine mammals from individual (eds.) Groundfish stock assessments length measurements (MS, available for the west coast in 1989 and from Marine Mammal Research recommended yield options for Unit, Fisheries Centre, 2044 Main 1990. Can. Tech. Rep. Fish. Aquat. Mall, University of British Sci. 1732. [41] Columbia, Vancouver, B.c. Canada V6T ILl). [29,51,55,70] Tyler, AV. and W.R. Crawford. 1991. Modeling of recruitment patterns in Trites, AW. and P.A Larkin. 1996. Changes Pacific cod (Gadus macrocephalus) in the abundance of Steller sea lions in Hecate Strait, British Columbia. (Eumetopias jubatus) in Alaska from Can. J. Fish. Aquat. Sci. 48:2240­ 1956 to 1992: how many were there? 2249. [42] (MS available from Fisheries Centre, University of British Columbia, 2044 Ulanowicz, R.E. 1986. Growth and Main Mall, Vancouver B.C. V6T development: ecosystems pheno­ lZ4). [27] menology. Springer-Verlag, New York, 203p. [15] Trumble, R.J. 1973. Distribution, relative Verbeek, N.AM. and R. Butler. 1989. abundance, and general biology of Feeding ecology of shoreline birds in selected underutilized fishery the Strait of Georgia. p. 74-80 In: resources ofthe eastern North Pacific Vermeer, K. and R.W. Butler (eds.) Ocean. M.S. Thesis, Univ. Washing­ The ecology and status ofmarine and ton, Seattle, Wash. 178p. [24] shoreline birds in the Strait of Georgia, British Columbia. Special Trumble, R.J., J.D. Neilson, W.R. Publication. Canadian Wildlife Bowering, and D.A McCaughran. Service, Ottawa. [73] 1993. Atlantic halibut (Hippoglossus hippoglossus) and Pacific halibut (H. Vermeer, K. 1981. Food and populations of stenolepis) and their North American surf sectors in British Columbia. fisheries. Can. Bull. Fish. Aquat. Sci. Wildfowl. 32: 107-116. [71,72] 227: 84p. [43,45]

113 Vermeer, K. and C.D. Levings. 1977. high seas driftnet fisheries of the Populations, biomass and food habits north Pacific Ocean. INPFC Bull. 53: of ducks on the Fraser Delta 285-296. [55] Intertidal Area, British Columbia. Wildfowl. 28: 49-60. [72] Walker, W.A., K.R. Goodrich, S. Leatherwood and RK. Stroud. 1984 Vermeer, K. and RC. Ydenberg. 1989. Population biology and ecology ofthe Feeding ecology of marine birds in Pacific white-sided dolphin, the Strait ofGeorgia. p. 62-73. In: K. Lagenorhynchus obliquidens, in the Vermeer and RW. Butler (eds.). northeastern Pacific. Part II: Biology 1989. The ecology and status of and geographical variation. marine and shoreline birds in the Administrative report LJ-84-34C of Strait of Georgia, British Columbia. the Southwest Fisheries Center, Special Publication. Can. Wildlife National Marine Fisheries Service, Service, Ottawa. [71, 72] P.O. Box 271, La Jolla, CA 92038. 39p. [54] Vermeer, K., 1. Robertson, RW. Campbell, G. Kaiser and M. Lemon. 1983. Walters, C. l, V. Christensen and D. Pauly. Distribution and densities of marine Stucturing dynamic models of birds on the Canadian West coast. exploited ecosystems from trophic Can. Wildlife Service, Vancouver. mass-balance assessments. Rev. Fish [56, 72] BioI. Fish. (in press). [82,86,90]

Vermeer, N.A.M. and R. Butler. 1981. Food Walters, C.J., R Hilborn, R.M. Peterman and populations of Surf Scoters in and M.l Stanley. 1978. Model for British Columbia. Wildfowl. 32: examining early ocean limitation of 107-116. Pacific salmon production. Can. J. Fish. Aquat. Sci. 35: 1303-1315. [68, Villavicencio, R.Z. and P. Muck. 1983. La 69] raeion de mantenimiento, la densidad de mantenimiento y la eficiencia de Ware, D.M. and G.A. MacFarlane. 1995. crecimiento de Engraulis ringens y Climate-induced changes in Pacific Sardinops sagax como una medida hake (Merluccius productus) de su potencia ecologica. BioI. Inst. abundance and pelagic community Mar. Peru. 7(4): 73-103. [41] interactions in the Vancouver Island upwelling system. Can Spec. Publ. Wahl, T.R., D.G. Ainley, A.H. Bennedict Fish. Aquat. Sci. 21: 509-521. [48, and A.R. DeGange 1989. 50] Associations between seabirds and water-masses in the northern Pacific Welch, D.W., L Margolis, M.A. Henderson Ocean in summer. Mar. BioI. 103: I­ and S. McKinnei. 1991. Evidence of ll. [31] attacks by the bathypelagic fish Anotopterus pharao Walker, W.A. and L.L. Jones. 1993. Food (Myctophiformes) on Pacific salmon habits of northern right whale (Oncorhynchus spp.). Can. J. Fish. dolphin, Pacific white-sided dolphin Aquat. Sci. 48:2403-2407. [25] and northern fur seal caught in the

114 Welschmeyer, N.A, S. Strom, R. Goericke, Marine Mammals. Vol. 3. The G. DiTullio, M. Belvin and W. sirenians and baleen whales. Petersen. 1993. Primary production Academic Press. [53] in the subarctic Pacific Ocean: Project SUPER. Prog. Oceanog. 32: Wood, C.C., K.S. Ketchen and RJ. 101-135. [17] Beamish. 1979. Population dynamics of spiny dogfish (Squalus acanthias) Welsh, H.E. 1968. Relationships between in British Columbia waters. J. Fish. assimilation efficiencies and growth Res. Board Can. 36: '647-656. [45] efficiencies for aquatic consumers. Ecology 49: 755-759. [20] Yamanaka, K.L. and L.J. Richards. 1992. Inshore rockfish. p. 221-266. In B.M. Westrheim, S.J. and lV. Tagart. 1984. Leaman (ed.) Groundfish stock Bathymetric distribution of Pacific assessments for the west coast of cod (Gadus macrocephalus) off Canada in 1991 and recommended British Columbia and Washington yield options for 1992. Can. Tech. State. Int. N. Pac. Fish. Comm. and Rep. Fish. Aquat. Sci. 1866. [67] recommended yield options for 1995. Can. Tech. Rep. Fish. Aquat. Zimmerman, ST. 1991. A history of marine Sci. 2069. 440p. [41] mammal stranding networks in Alaska, with notes on the distribution Westrheim, SJ. and R.P. Foucher. 1987. of the most commonly stranded Stock assessment of Pacific cod cetacean species, 1975-1987. p. 43­ (Gadus macrocephalus) in Georgia 53. In: Marine Mammal Strandings in and Juan de Fuca Straits. Can. MS the United States; Proceedings of the Rep. Fish. Aquat. Sci. 1905. 84p. Marine Mammal Stranding Workshop [67] (2nd), held on December 3-5, 1987, Miami, Florida. [30] Westrheim, SJ. and W.R. Harling. 1983. Principal prey species and periodicity of their incidence in stomachs oftrawl-caught cod (Gadus macrocephalus), rock sole (Lepidopsetta bilineata), and petrale sole (Eopsetta jordani) landed in British Columbia, 1950-1980. Can. MS Rep. Fish. Aquat. Sci. No. 1681. 38p. [42]

Wilke, F., T. Taniwaki and N. Kuroda. 1953. Phocoenoides and Lagenorhynchus in Japan, with notes on hunting. 1 Mamm. 34: 488-497. [30]

Wolman, AA 1985. Gray whale Eschrichtius robustus (Lilljeborg, 1981). p. 67-89. In: S.H. Ridgway and R. Harrison (eds.) Handbook of

115 Appendix 1

Table A Biological and population statistics of marine mammals, Alaska Gyre (Summer).

Species Mean weight (kg) Popu- Sex ratio Area Ration (kg/day) Q/B 3 2 (males) (females) lation (females) (10 km ) (males) (females) (per year) Northern fur seals 30 25 130000 0.8 70 1.5 1.3 20.7 Steller sea lions 210 186 0 0.6 70 7.2 6.5 0.0 N. elephant seals 412 330 4000 0 70 12.4 10.4 9.2 PINNIPEDS 16.9 Killer whale (resid.) 2587 1973 238 0.64 70 84.6 84.1 14.0 Dall's porpoise 63 61 21200 0.5 4420 2.8 2.7 16.0 Spenn whales 27400 26938 2000 0 612 822 351 4.7 TOOTHED WHALES 5.4 Baird's beaked whales 3134 3833 10000 0.4 10600 62.6 73.6 7.4 Stejneger's b. whale 497 511 1000 0.5 5300 14.4 14.7 10.5 Cuvier's b. whale 961 893 1000 0.5 10000 24.3 22.9 9.3 BEAKED WHALES 7.5 Blue whale 95347 110126 1700 0.5 5300 3814 4405 14.6 Fin whale 51361 59819 17000 0.5 5300 2055 2393 14.6 Sei whale 16235 17387 14000 0.5 5300 649 696 14.6 BALEEN WHALES 14.6 KILLER W. (trans.) 2587 1974 44 0.64 70 73.0 73.0 12.1

Table B Biological and population statistics of marine mammals, Alaska Gyre (Winter).

Species Mean weight (kg) Popu- Sex ratio Area Ration (kg/day) Q/B 3 2 (males) (females) lation (females) (10 km ) (males) (females) (per year) Northern fur seals 30 25 5000 0.8 70 1.5 1.3 20.7 Steller sea lions 210 186 5000 0.6 70 7.2 6.5 13.0 N. elephant seals 412 330 4000 0 70 12.4 10.4 9.2 PINNIPEDS 11.1 Killer whale (resid.) 2587 1973 238 0.64 70 84.6 84.1 14.0 Dall's porpoise 63 61 21200 0.5 4420 2.8 2.7 16.0 Sperm whales 10098 26938 0 0 612 160 351 0.0 TOOTHED WHALES 14.1 Baird's beaked whales 3134 3833 0 0.4 10600 62.6 73.6 '0.0 Stejneger's b. whale 497 511 1000 0.5 5300 14.4 14.7 10.5 Cuvier's b. whale 961 893 1000 0.5 10000 24.3 22.9 9.3 BEAKED WHALES 9.9 Blue whale 95347 110126 1700 0.5 5300 963 1080 3.6 Fin whale 51361 59819 17000 0.5 5300 587 663 4.1 Sei whale 16235 17387 14000 0.5233 5300 234 247 5.2 BALEEN WHALES 4.2 KILLER W. (trans.) 2587 1974 44 0.64 70 73.0 73.0 12.1

116 Table C Diet composition of marine mammals in the Alaska gyre (Summer).

diet compositions Species benthic euphaus. Cope- small large small meso salmon large org. pods squids squids pel. pel. pel. northern fur seals 0.00 0.00 0.00 0.78 0.00 0.00 0.00 0.11 0.11 Steller sea lions 0.00 0.00 0.00 0.30 0.20 0.10 0.00 0.00 DAD N elephant seals 0.00 0.00 0.00 DAD 0.20 0.00 0.25 0.00 0.15 pinnipeds 0.000 0.000 0.000 0.713 0.035 0.000 0.044 0.091 0.117 Killer whale (resid.) 0.00 0.00 0.00 0.00 0.00 0.10 0.00 0.80 0.10 Dall's porpoise 0.00 0.00 0.00 0.20 0.00 0.10 0.70 0.00 0.00 Sperm Whales 0.00 0.00 0.00 0.05 0.80 0.00 0.00 0.00 0.15 toothed whales 0.000 0.000 0.000 0.041 0.634 0.021 0.006 0.159 0.139 Baird's beaked whale 0.00 0.00 0.00 0.30 0.35 0.10 0.00 0.00 0.25 Stejneger's beaked w 0.00 0.00 0.00 0.50 0045 0.00 0.00 0.00 0.05 Cuvier's beaked w 0.00 0.00 0.00 0.35 0.35 0.00 0.15 0.00 0.15 beaked whales 0.000 0.000 0.000 0.309 0.354 0.093 0.005 0.000 0.239 blue whale 0.00 0.95 0.05 0.00 0.00 0.00 0.00 0.00 0.00 fin whale 0.00 0.75 0.20 0.00 0.00 0.05 0.00 0.00 0.00 sei whale 0.00 0.10 0.80 0.05 0.00 0.05 0.00 0.00 0.00 baleen whales 0.000 0.663 0.285 0.009 0.000 0.044 0.000 0.000 0.000

Table D Diet composition of marine mammals in the Alaska gyre (Winter).

diet compositions Species benthic euphaus. Cope- small large small meso salmon large org. pods squids squids pel. pel. pel. northern fur seals 0.00 0.00 0.00 0.15 0.15 0.25 0.15 0.00 0.30 Steller sea lions 0.00 0.00 0.00 0.15 0.00 0.20 0.00 0.00 0.65 N elephant seals 0.00 0.00 0.00 DAD 0.20 0.00 0.25 0.00 0.15 pinnipeds 0.000 0.000 0.000 0.274 0.112 0.105 0.137 0.000 0.371 Killer whale (resid.) 0.00 0.00 0.00 0.00 0.00 0.20 0.00 0.60 0.20 Dall's porpoise 0.00 0.00 0.00 0.20 0.00 0.10 0.70 0.00 0.00 Sperm Whales 0.00 0.00 0.00 0.15 0.60 0.05 0.05 0.00 0.15 toothed whales 0.000 0.000 0.000 0.009 0.000 0.196 0.030 0.574 0.191 Baird's beaked whale 0.00 0.00 0.00 0.35 0.30 0.10 0.10 0.00 0.15 Stejneger's beaked w 0.00 0.00 0.00 0.50 0045 0.00 0.00 0.00 0.05 Cuvier's beaked w 0.00 0.00 0.00 0.35 0.35 0.00 0.15 0.00 0.15 beaked whales 0.000 0.000 0.000 0.431 0.404 0.000 0.069 0.000 0.096 blue whale 0.00 0.90 0.10 0.00 0.00 0.00 0.00 0.00 0.00 fin whale 0.00 0.60 0.20 0.05 0.00 0.05 0.05 0.00 0.05 sei whale 0.00 0.10 0.70 0.05 0.00 0.05 0.05 0.00 0.05 baleen whales 0.000 0.526 0.296 0.044 0.000 0.044 0.044 0.000 0.044

117 Table E Diet matrix used for Alaska gyre model.

Predator Prey 3 4 5 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 I Phytoplankton .75 .915 .53 .85 - .30 ------2 Bacteria .25 - -- .10 ------3 Microzoopl. - .40 - - .10 ------4 Sm. herb. - -- .05 .63 -- .03 -- .6 .75 - .05 - - .285 - -- 5 Salps ------6 Cam. zoopl. ------.10 - .05 ------7 Jellies ------.98 ------8 Krill - -- - .20 --- -- .1 .10 ---- .662 - .06 ...... - ...... 9 Squids - -- - .05 - .75 .75 .02 .95 - -- .75 .748 .678 .009 .663 - .45 00 10 Crustaceans -- - - .05 - .15 .20 - .04 .3 - - .03 --- -- .01 II Pink ------.050 - .021 .040 -- - - 12 Sockeye ------.050 - .025 .030 - - - - 13 Chum ------.050 - .020 .040 - -- - 14 Steelhead ------.050 - .025 .046 - - -- 15 Mesopelagics - -- - .02 ------.05 .200 .08 .044 .006 - .005 - - 16 Sm. pelagics - - - - .05 - .10 .02 - .01 - - .394 .04 - .021 .044 .093 - .48 17 Sharks ------.001 ------18 Large fish ------.200 - .117 .139 - .239 - - 19 Pinnipeds ------.005 - - --- .025 - 20 Tth. whales ------.125 - 21 Bal. whales ------.100 - 22 Beak. whales ------23 Orca ------24 Mar. birds ------25 Detritus - .085 .07 .10 - .50 ------Import ------.750 - Table F Population statistics of marine mammals, southern shelf ofD.C. (Summer).

Mean body Popu- Sex Consumption Q/B l Species weight (kg) lation ratio (kg'daf ) (year"l) males females females male female annual Northern fur seals 30 25 61000 0.8 1.5 I.3 20.7 Steller sea lions 210 186 9400 0.6 7.2 6.5 13.0 N. elephant seals 412 3308 2000 0 12.4 10.4 9.2 California sea lions 86 585 0 0 3.5 2.6 0.0 Harbour seals 64 56 20000 0.5 2.8 2.5 16.1 Pinnipeds 15.3 White-sided dolphin 85 73 2000 0.5 3.5 3.1 15.2 Killer whale (resid) 2587 1974 200 0.64 85 84 14.0 Dall's porpoise 63 619 1000 0.5 2.8 2.7 16.0 Harbour porpoise 29 33 1000 0.5 1.5 1.6 18.4 N. right whale dolphin 142 682 100 0.5 5.3 2.9 14.2 Toothed whales 15.6 Humpback whale 28323 32493 100 0.5 365 407 4.6 Gray whale 15920 16435 1167 0.5 637 657 14.6 Minke 6121 7011 100 0.5 107 119 6.3 Baleen whales 13.0 Killer whales (trans) 2587 1974 34 0.64 73 73 12.1

Table G Population statistics of marine mammals, southern shelf of D.C. (Winter).

Mean body Popu- Sex Consumption Q/B Species weight. (kg) lation ratio (kg'day"l) (year"l) males females females male female annual Northern fur seals 30 25 51000 0.8 1.5 I.3 20.7 Steller sea lions 210 186 9400 0.6 7.2 6.5 13.0 N. elephant seals 412 3308 2000 0 12.4 10.4 9.2 California sea lions 86 585 3500 0 3.5 2.6 11.0 Harbour seals 64 56 20000 0.5 2.8 2.5 16.1 Pinnipeds 14.8 White-sided dolphin 85 73 2000 0.5 3.5 3.1 15.2 Killer whale (resid) 2587 1974 290 0.64 85 84 14.0 Dall's porpoise 63 619 1000 0.5 2.8 2.7 16.0 Harbour porpoise 29 33 1000 0.5 1.5 1.6 18.4 N. right whale dolphin 142 682 100 0.5 5.3 2.9 14.2 Toothed whales 15.3 Humpback whale 28323 32493 100 0.5 365 407 4.6 Gray whale 15920 16435 585 0.5 637 657 5.3 Minke 6121 7011 300 0.5 107 119 6.3 Baleen whales 5.1 Killer whales (trans) 2587 1974 85 0.64 73 73 12.1

119 Table H Diet composition of marine mammals, southern shelfof H.C. (Summer).

Species octopus krill cope- bival- poly- amphi- small large sand- smelts pods ves chaetes pods squids squids lance Northern fur seals ------.14 - .03 .01 Steller sea lions .02 - - - - - .01 - .08 .01 N. elephant seals ------.40 -- - Cal. sea lions .01 - - - - - .02 - .05 .02 Harbour seals ------.02 - .02 .02 Pinnipeds .007 - - - - - .109 - .040 .011 W.-sided dolphin ------.10 - .30 - Killer whale (resid) ------Dall's porpoise ------.20 - .30 - Harbour porpoise ------.20 - .40 - N. right whale dolph. ------.10 - .40 - Toothed whales ------.055 - .127 - Humpback whale - .80 .10 ------Gray whale - - - .05 .05 .90 - - - - Minke - .30 .30 - - - - - .20 - Baleen whales - .043 .009 .047 .047 .844 - - -3 -

Table H (cont.)

Species herring other hake rock- sable cod coho, pink, chum pollock var. sharks dup. fish fish chinook sockeye pelagics Northern fur seals .52 .02 .03 .07 .04 .06 .06 .02 Steller sea lions .16 .01 .Il .12 .09 .09 .26 .05 N. elephant seals .10 .10 .10 .10 .20 Cal. sea lions .16 .33 .06 .06 .05 .05 .17 .04 Harbour seals .15 .65 .02 .02 .08 .03 Pinnipeds .239 .006 .170 .073 .027 .040 .049 .049 .126 .023 .030 W.-sided dolphin .40 .10 .10 Killer whale (resid) .05 .65 .15 .15 Dall's porpoise .40 .10 Harbour porpoise .40 N. right whale dolp. .50 Toothed whales .192 .023 .389 .113 .010 .090 Humpback whale .10 Gray whale Minke .20 Baleen whales .008 -0

120 Table I Diet composition of marine mammals, southern shelf of D.C. (Winter).

Species octopus krill cope- bivalves poly- amphi- small large sand- smelts pods chaetes pods squids squids lance Northern fur seals ------.\4 - .03 .0\ Steller sea lions .04 -- - - - .02 - .04 .02 N. elephant seals ------.40 -- - Cal. sea lions .02 -- --- .02 - .02 .02 Harbour seals ------.02 -- .0\ Pinnipeds 0\4 ----- .106 - .022 .0\2 W.-sided dolphin ------.10 - .30 - Killer whale (resid) ------Dall's porpoise ------.20 - .30 - Harbour porpoise ------.20 - .40 - N. right whale dolph. ------.10 - .40 - Toothed whales ------.044 - .\00 - Humpback whale - .80 .10 ------Gray whale --- .05 .05 .90 ---- Minke - .30 .30 ----- .20 - Baleen whales . .3\6 .074 .026 .026 .47\ -- .026 -

Table I (cont.)

Species herring other. hake rock- sable cod coho, pink, chum pollock var. sharks dup. fish fish chinook sockey pelagics Northern fur seals .52 .02 .03 .07 .04 - .06 .06 .02 - - Steller sea lions .13 - - .05 - .22 .06 .06 .35 .02 - N. elephant seals - - .10 .10 .10 --- .10 - .20 Cal. sea lions .37 - .10 .03 - .\1 .04 .04 .22 .04 . Harbour seals .60 - .20 -- - .02 .02 .09 .05 - Pinnipeds .321 .005 .071 0050 .025 .080 .038 .0038 .\68 .020 .030 W.-sided dolphin .40 -- .10 -- . .\0 . - - Killer whale (resid) .03 - ---- .76 .10 -- .\0 Dall's porpoise .40 ------.10 - - Harbour porpoise .40 ------N. right whale dolph. .50 ------Toothed whales .\5\ -- .018 -0 - .520 .089 .008 - .070 Humpback whale .10 . ------Gray whale ------Minke .20 ------Baleen whales .06\ ------

121 3 Table J Diet matrix used for the southern B.C. shelf modeI. )

Predator Prey ] 2 3 4 5 6 7 8 9 ]0 ]] 12 13 ]4 ]5 ]6 ]7 ]8 19 20 21 22 23 24 25 I Transient orcas ------2 Odontocetae .20 ------3 Pinnipeds .75 ------4 Mysticetae .05 ------5 Seabirds ------. -- 6 Spiny dogfish - .602 .278 - .016 .05 ------7 Pacific halibut ------8 Pacific hake - - .170 - - .02 - .010 .050 .16 .05 ------9 Pacific cod - .023 .113 - .06 .18 .005 .090 - .05 - - - - N ------N 10 Juv. sablefish - - .027 -- -- - .010 -- . ------. --- - - II Adult sablefish ------12 Herring/sm. pel. - .320 .296 .012 .424 .40 .70 .050 .250 .41 .40 ------13 Cam. jellies - - -- .005 .05 - - - .04 .03 - .050 ------14 Decapods - - - - .040 .08 .]0 .014 .050 .05 .02 -- .03 ------15 Shrimps - .055 .116 - .102 .05 .02 - .050 .04 .39 -- .10 ------16 Polychaetes - - - .045 - - - - .099 - - - .050 - .3 - .29 ------17 Sea stars - - - - .001 ------.05 ------]8 Sea urchins ------.050 ------]9 Brittle stars ------20 Bivalves - - - .047 ------.10 -- .05 ------. 21 Amphipods - - - .844 --- - .100 .05 .02 - - .02 ------22 Euphausiids -- - .043 .172 .20 - .355 .100 .25 .02 .01 ------23 Chaetognaths - - -- .079 .03 - --- - .01 .080 - --- - .05 ------24 Salps -- - -- .06 - - -- .02 - .050 ------1 1 - - 25 Copepods - - - .009 .161 - - - .100 - - .98 .670 - -- -- .05 ------26 Phytoplankton ------. ------. - - - .9 .9 27 Detritus ------.050 - . - .099 .84 .7 I .61 1 .90 I 1 - - .1 .1 ImPOrt ------.565 ------a) due to rounding errors, the diet composition ofa few groups add up to 0.999, and not to 1.000; this does not affect computations Table K Means seasonal landing of the fisheries, southern B.C. shelf (in t·km-2 per 6 month season; based on DFO data for the years 1985- 1995).

FUNCTIONAL GROUP: Species Summer Winter BIVALVES: Butter clam 0.03 2.42 Geoduck 969.4 682.54 Horse clam 44.04 3.08 Manila clam 58.51 374.08 Mixed clam 1.68 15.60 Mussel 0 0.10 Native littleneck clam 3.43 12.33 Scallop 0.36 0.21 MISe. BENTHOS: Abalone 0.48 1.00 Gooseneck barnacle 19.4 8.31 ECHINODERMS: Sea cucumber 72.82 139.85 Red sea urchin 57.58 356.64 Green sea urchin 0 9.90 Purple sea urchin 0 0.18 OCTOPUS: Octopus 9.64 7.08 SQlJIDS: Squid 72.38 2.87 ZOOPLANKTON: Euphausiids 0 0.44 COOS/BLACK COOS: Lingcod 2278.29 553.76 Pacific cod (grey cod) 1364.01 1096.46 Sablefish (black cod) 1557.17 481.1 0 SHRIMPS/CRABS: Dungeness crab 202.81 101.53 Red rock crab 1.98 0 King crab 0 0.15 Prawn 57.76 24.46 Crayfish 0.02 0 Shrimp 1517.76 375.32 LARGE PELAGICS: Hake 19500 2125.79 Tuna 50.26 30.53 SMALL PELAGICS: Herrings 0.16 0.10 Anchovy 23.76 4.24 Mackerel 5.12 0.30 MISe. DEMERSALS: Halibut: 626.25 77.22 Brill sole 204.77 246.27 Butter sole 0.04 0.02 Dover sole 442.74 542.39 Lemon sole 95.18 28.46 Mixed sole 37.21 21.21 Rex sole 6.64 5.56 Rock sole 244.18 59.52 Flounder 5.66 30.57 Turbot 829.95 140.11 Skate 31.72 13.74 Hagfish 66.53 64.33 Pollock 69.27 76.54 Sturgeon 0.45 0.31 Greenlings. etc. 0.08 0.01 ROCKFISHES: Perch (silver) 0.38 0.47 Pacific ocean perch 857.7 627.62 Greenies (tellow tail) 842.26 1734.45 Idiot rockfish 33.46 37.81 Red snapper (yellow eye) 314.49 135.72 Reedi (yellow mouth) 393.25 401.25 Rockfish 3218.23 3079.39 SHARKS: Dogfish: 1071.3 1025.01 Other sharks 0.72 1.19 TRANSIENT SALM.: Steelhead 3.45 0.29 Chum 458.41 3773.90 Pink 1814.98 14.91 Sockeye 4013.71 1.79 RESIDENT SALM.: Chinook 2431.58 20.20 Coho 4389.53 44.13 OTHER: Non-food catch 83.57 71.34 Other fish 1.07 0.11 Roe on kelp 34.43 0.63 Roe herring 0 6056.37 Salmon roe 0.86 0.47

123 2 Tabl~ L Standing stock of macrobenthos (t·km- ) of Strait ofGeorgia, by depth range.

Depth range (m): >400 300-400 200-300 100-200 50-100 20-50 <200 Intertidal Surface (km; total: 8635) 330 1000 1570 2130 1560 800 622.5 622.5 Annelids 7.3 Polychaets 11.3 15 29 9.0 222 58 144 Amphipods 0.2 0.50 0.80 0.60 0.20 14 Copepods 1.0 0.30 Cumaceans 0.20 Cirripeds 306 Isopods 7.3 1.2 0.20 27 Pelecypod (bivalves) 7.0 0.30 310 118 347 177 21 942 Scaphopods 0.85 2.7 0.50 0.40 Gastropods 1.4 0.70 1.4 1.8 2.7 3.8 3.2 76 Amphineura 0.2 2.0 0.65 0.37 0.7 186 Arthropods 17 0.20 Ophiurids 0.70 0.85 1.2 11 12 1.4 0.10 Echinoids 29 25 271 116 0.20 Holothurian 36 48 156 63 56 1.5 3.2 Nemertines 1.7 4.3 1.7 5.0 Sipunculoids 0.30 0.5 2.0 0.60 Echiuroidea 65 0.7 Porifers (sponges) 0.20 4.0 8.3 0.50 1162 Cnidaria (anthozoa) 2.3 5.4 0.30 0.64 16 58 Unidentified benthos 4.3 1.5 0.60 Asteroida 0.30 0.20 23 Decapods 1.5 12 1.1 13 124 All macrobenthos 104 99 430 136 650 446 177 916

124 Table M Diet Composition of Marine Mammals in the Strait of Georgia.

Mammal species DC, per area (%) DC, whole Food cons. Estuaries / non-estuaries Strait (%) (t'year"l) Killer whales (resident) Small squid - / - 5.0 72.4 Large squid - / - 5.0 72.4 Small pelagics - / - 10.0 144.7 Miscellaneous fishes - / - 40.0 578.9 Higher invertebrates - / - 40.0 578.9 Killer whales (transient) Harbour seal -/ - 56.6 134.4 Porpoise -/ - 2\.5 51.1 Whale -/ - 21.5 51.1 Bird -/ - 0.3 0.7 Squid - / - 0.04 0.1 Dall's porpoises Benthic invertebrates - / - 5.0 55.9 Small squid - / - 30.0 335.6 Large squid - / - 10.0 II\.9 Small pelagics - / - 20.0 223.7 Mesopelagics - / - 20.0 223.7 Miscellaneous fishes -/ - 15.0 167.8 Harbour porpoises Benthic invertebrates - / - 5.0 14.3 Small squid - / - 10.0 28.5 Large squid - / - 10.0 28.5 Small pelagics -/ - 30.0 85.63 Miscellaneous fishes - / - 45.0 128.3 Harbour seals (1988 data) Salmonids 9/4 4.5 448.8 Gadoids 50/65 63.5 6,303.6 Pacific herring 19/12 12.7 1,264.1 Sculpins 5/ I 1.4 140.4 Flatfishes 5/ I 0.5 51.4 Surfperches 4/3 3.1 308.3 Hexagrammids 0/1 0.9 89.1 Plainfin midshipman 3/2 2.1 209.0 Sandlance 0/2 1.8 178.2 Cephalopods 1 /2 1.9 188.4 Other 3/5 4.8 476.2 Unidentified 1 /3 2.8 277.5

125 Table N Diet Composition of Marine Birds in the Strait of Georgia.

Bird species DC Food cons. Bird Species DC Food cons. l l % (t·year· ) % (t·year· ) Glaucous-winged gull White-winged scoters Refuse 65.0 1852.3 clams 80.0 135.8 Bivalves 20.0 570.0 snail 8.0 13.6 Pacific herrings 15.0 427.5 errant polychaetes 4.0 6.8 Double-crested cormorant barnacles 8.0 13.6 penpoint gunnel 36.0 250.8 Other diving ducks (=oldsquaw) shiner perch 21.0 146.3 bivalves 64.0 1376.9 crested gunnel 16.0 111.5 snail 3.0 64.5 snake prickleback 10.0 69.7 crustaceans 33.0 710.0 others 17.0 118.4 Pigeon guillemot Pelagic cormorant benthic fish 70.0 34.3 crescent gunnel 37.0 282.8 herring 3.3 1.6 Pacific sandlance 19.0 145.2 smaltt polychaetes 3.3 1.6 Pacific staghorn sculpin 13.0 99.4 lingcod 3.3 1.6 penpoint gunnel 11.0 84.1 cod 3.3 1.6 others 20.0 152.9 sculpins 3.3 1.6 Brandt's cormorants flatfish 3.3 1.6 herring 26.0 88.5 rockfish 3.3 1.6 giant marbled sculpin 26.3 89.4 squid 3.3 1.6 lithodid crab 8.8 29.9 shrimp 3.3 1.6 shrimp 39.0 133.0 Other auks (=Pigeon guillemot) Loons (Arctic & others) Benthic fish 70.0 132.9 herring 100.0 471.7 herring 3.3 6.3 Grebes (West. & other) smalt 3.3 6.3 herring 70.0 1398.8 lingcod 3.3 6.3 shiner perch 20.0 399.7 cod 3.3 6.3 algae 4.0 79.9 sculpins 3.3 6.3 eelgrass 4.0 79.9 flatfish 3.3 6.3 snail 2.0 40.0 rockfish 3.3 6.3 Su rf scoters squid 3.3 6.3 blue mussels 85.0 961.2 shrimp 3.3 6.3 snail 5.0 56.5 Great blue herons errant polychaetes 5.0 56.5 gunnels 48.4 95.2 barnacles 5.0 56.6 staghom sculpin 24.2 47.6 shiner perch 14.1 27.7 three-spined stickleback 9.2 18.1 bay pipefish 2.3 4.5 chum salmon 1.7 3.3

126 Table 0 Diet matrix used for the Strait ofGeorgia model.

Predator

3 Prey 1 2 3 4 5 6 7 8 9 10 11 12 13 14 ) 1 Mammals (res.) ------1 - 2 Large pelagics .010 -- .100 -. ------3 Small pelagics .300 .200 - .200 .100 ------.220 -- 4 Hake .300 .100 - .070 .050 ------5 Misc. demersals .101 .050 - .070 ------.150 - - 6 Jellies - .010 - -- .030 ------7 Lg. macrobenthos .009 - - - .200 - .050 -- - - .200 -- 8 Sm. macrobenthos .040 - .100 - .200 - .300 .050 --- .190 - - 9 cam. zooplankton .010 .450 .100 .350 .400 .260 ------10 Herb. zooplankton - .190 .800 .210 .050 .710 -- .500 - - .060 - - 11 Primary producers ------.250 .900 - .010 .050 - 12 Birds ------.050 - 13 Transient orcas ------.050 - 14 Salmon .230 ------15 Detritus - - --- .650 .950 .250 .100 - .170 - - a) Salmon is considered not to eat within the Strait ofGeorgia (see text).

127 Appendix 2

Workshop Schedule

November 6-10,1995 Fisheries Centre, University of British Columbia

Sunday 5 Arrival, and settling into Gage Court Monday 6 Workshop Day 1 at Ralf Yorque Room, Fisheries Centre 0915 - 0930 Welcome.....Tony Pitcher, Director, Fisheries Centre 0930 - 0945 Introductory remarks Peter Larkin 0945 - 1000 About this workshop Daniel Pauly 1000 -1045 Coffee break 1045 - 1200 Lecture 1 The Ecopath Approach: Theory and Applications. Villy Christensen 1200 - 1400 Sandwich lunch in Ralf Yorque Room 1400 - 1700 Workshop Session 1: Definition ofperiod (seasons/years/decades) and ofareas to be covered by models. Moderator: D. Pauly. 1700 - 1830 Refreshment in Ralf Yorque Room Tuesday 7 Workshop Day 2 0900 - 1000 Lecture 2 Variation ofNorth Pacific climate and marine ecosystems. Jeffrey Polovina 1000 - 1045 Coffee break 1045 - 1200 Workshop Session 2: Definition offunctional groups ("boxes") to be included in models and assignment of boxes to participants. Moderator: V Christensen 1200 - 1400 Sandwich lunch in Ralf Yorque Room 1400 - 1700 Workshop Session 3: Participants assemble key parameter estimates (biomass, mortality, etc. ) for their group Moderator: D. Pauly Wednesday 8 Workshop Day 3 0900 - 1000 Student presentation: A trophic model ofthe Strait of Georgia. Moderator: Rik. Buchvorth 1000 - 1045 Coffee break 1045 - 1200 Workshop Session 4: Assembling the diet matrix Moderator: D. Pauly 1200 - 1400 Sandwich lunch in Ralf Yorque Room 1400 - 1700 Workshop Session 5: Data entry and balancing ofa preliminary model. Moderator: V Christensen Thursday 9 Workshop Day 4 0900 - 1000 Group work 1000 - 1045 Coffee Break 1045 - 1200 Workshop Session 6: Construction and discussion ofmodel alternatives by participants. Moderator: D. Pauly

128 1200 - 1400 Sandwich lunch in Ralf Yorgue Room 1400 - 1700 Workshop Session 7: Discussion offlow networks and ancillary statistics ofbalanced models. Moderator: V Christensen 1900 Buffet Dinner at Green College - UBC (Small Dining Room, Graham House) Friday 10 Workshop Day 5 0900 - 1000 Lecture 3 Comparative studies on eastern boundary current (upwelling) ecosystems. A. Jarre-Teichmann 1000 - 1030 Coffee Break 1030 - 1200 Workshop Session 8: Wrapping up - what has been learnt from all this. Moderator: V Christensen 1200 End of Workshop (Lunch at Trekker's Restaurant) 1400 - 1500 Fisheries Centre Seminar at Ralf Yorque Room "Why do fish stocks collapse? The case ofcod in Eastern Canada" Ransom A. Myers, Department of Fisheries and Oceans, S1. John's Newfoundland.

129 Appendix 3

List ofParticipants

Dr. Lee Alverson Dr. Patricia Livingston Natural Resources Consultants, Inc. Alaska Fisheries Science Centre 4055 21st Avenue West 7600 Sand Point Way N.E. Seattle, WA 98199, USA Seattle, WA 98155, USA Tel: (206) 285-3480 Tel. (206) 526-4242 Fax: (206)283-8263 e-mail:[email protected]

Dr. Mary Arai Dr. Daniel Pauly Department ofBiological Sciences Fisheries Centre University ofCalgary University ofBritish Columbia Calgary, AB T2N 4H5, CANADA Vancouver, B.C. V6T lZ4, CANADA Tel: (403) 220-5281 Tel: (604) 822-1201 Fax: (403) 289-9311 Fax: (604) 822-8934 e-mail: [email protected] e-mail: [email protected]

Dr. Villy Christensen Dr. Tony Pitcher ICLARM,North Sea Centre Fisheries Centre DK-9850, Hirtshals University ofBritish Columbia DENMARK Vancouver, B.C. V6T lZ4, CANADA Tel: +45 9894 4622 Tel: (604) 822-2731 Fax: +45 3396 3260 Fax: (604) 822-8934 e: mail: [email protected] e-mail: [email protected]

Dr. Astrid Jarre-Teichmann Dr. Jeffrey Polovina Institute ofMarine Science National Marine Fisheries Service Department ofFisheries Biology 2570 Dole Street Duestembrooker Weg 20 Honolulu, Hawaii 96822-2396 24105 Kiel, GERMANY USA Tel: (49-431) 597-3921 Tel: (808) 943-1221 Fax: (49-431) 565-876 Fax: (808) 943-1290 e-mail: [email protected] e-mail:[email protected]

Dr. Peter Larkin Dr. Jenny Purcell The North Pacific Universities Marine Hom Point Env. Lab. Mammal Reearch Consortium P.O. Box 775 University ofBritish Columbia Cambridge, Maryland 21613, USA Vancouver, B.C. V6T lZ4 CANADA e-mai: [email protected]

130 Dr. Andrew Trites Dr. Dan Ware The North Pacific Universities Marine DFO, Pacific Biological Station Mammal Research Consortium 3190 Hammond Bay Road University ofBritish Columbia Nanaimo, B.C. V9R 5K6, CANADA Vancouver, B.C. V6T lZ4 CANADA Tel: (604) 756-7199 Tel: (604) 822-8181 Fax: (604) 756-7053 Faz: (604) 822-8180 e-mail: [email protected]

Dr. Carl Walters Fisheries Centre University ofBritish Columbia Vancouver, BC V6T lZ4, CANADA Tel: (604) 822-6320 Fax. (604) 822-8934 e-mail: [email protected].

Students participants

Mailing address for all: Fisheries Centre 2204 Main Mall University ofBritish Columbia Vancouver, BC CANADA V6T lZ4

Name e-mail address Ms. Eny Buchary [email protected] Mr. Rik Buckworth [email protected] Ms. Alida Bundy [email protected] Ms. Sylvie Guenette [email protected] Ms. Kathy Heise [email protected] Mr. Narriman Jiddawi [email protected] Mr. John Kelson Mr. Steven Mackinson [email protected] Ms. Claudia Octeau [email protected] Mr. Judson Venier [email protected] Mr. Yoshihiko Wada [email protected]

Workshop Coordinator: Ms. Ying Chuenpagdee Fisheries Centre 2204 Main Mall Vancouver, B.y. V6T lZ4, CANADA Tel. (604) 822-0618 Fax (604) 822-8934 e-mail: [email protected]

131 ISSN 1198-6727

FISHERIES CENTRE RESEARCH REpORT SERIES

Commercial Whaling - the Issues Reconsidered June 10-11, 1993 Fisheries Centre Research Reports 1993, Volume 1, Number 1 36pp

Decision Making by Commercial Fishermen: a missing component in fisheries management? November 25-26, 1993 Fisheries Centre Research Reports 1993, Volume 1, Number 2 75pp

Bycatch in Fisheries and their Impact on the Ecosystem October 13-14, 1994 Fisheries Centre Research Reports 1994, Volume 2, Number 1 86pp

Graduate Student Symposium on Fish Population Dynamics and Management April 22-23, 1995 Fisheries Centre Research Reports 1995, Volume 3, Number 1 30pp

Harvesting Krill: Ecological impact, assessment, products and markets November 14-16,1995 Fisheries Centre Research Reports 1995, Volume 3, Number 3 82pp

Mass-Balance Models of North-eastern Pacific Ecosystems November 6-10, 1995 Fisheries Centre Research Reports 1996, Volume 4, Number 1 131pp

Reinventing Fisheries Management February 21-24, 1996 Fisheries Centre Research Reports 1996, Volume 4, Number 2 84pp

Copies of any of these research reports may be obtained at a cost of $20.00, which includes surface mail. Payment may be made by Credit Card. Please contact:

Events Officer Fisheries Centre University of British Columbia 2204 Main Mall Vancouver V6T 1Z4 Canada

Phone: 604-822-2731 Fax: 604-822-8934 E-mail: [email protected] Web Site: http://fisheries.com